To see the other types of publications on this topic, follow the link: Fashion-MNIST dataset.
Journal articles on the topic 'Fashion-MNIST dataset'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Fashion-MNIST dataset.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
ГНАТУШЕНКО, В. В., Т. М. ФЕНЕНКО та О. Л. ДОРОШ. "РЕЗУЛЬТАТИ НАЛАШТУВАННЯ ПАРАМЕТРІВ НЕЙРОННИХ ГЛИБОКИХ МЕРЕЖ ЩОДО РОЗПІЗНАВАННЯ FASHION MNIST DATASET". Applied Questions of Mathematical Modeling 5, № 2 (2023): 19–26. http://dx.doi.org/10.32782/mathematical-modelling/2022-5-2-2.
Full textAbstract:
Проведено дослідження моделей згорткової нейронної мережі (Convolutional neural network – CNN) з метою підвищення точності розпізнавання Fashion MNIST DATASET. З огляду відомо, що розпізнавання елементів одягу набору Fashion MNIST є більш складним ніж розпізнавання набору рукопису цифр MNIST. Набор одягу Fashion-MNIST рекомендовано для досліджень різних архітектур нейронних мереж. Найкращі результати якості розпізнавання Fashion MNIST DATASET отримано за згортковою нейронною мережею. В даній роботі було метою покращити точність розпізнавання Fashion MNIST DATASET за рахунок дослідження різних архітектур CNN та їх параметрів. Обрано дві архітектури послідовної згорткової нейронної мережі з тих, у яких точність розпізнавання Fashion MNIST DATASET більше ніж 93%. Проведено дослідження їх архітектур та параметрів. Моделі відповідають визначенню нейронних глибоких мереж та мають різну кількість шарів. В дослідженнях моделей показано вплив параметрів batch_size, validation_split, validation_data на точність розпізнавання, а також варіанти розташування шару BatchNormalization та шару активації; вплив параметра “filters” для згорткового шару. Крім того, було використано два варіанти вибору валідаційної вибірки: перший – з набору даних для навчання (20%), а другий – набор даних тестування. При розрахунках число епох навчання дорівнювало 20. В процесі навчання вирішувалось питання не допустити перенавчання за допомогою аналізу функції втрат. Використано бібліотеки TensorFlow, Keras, мову програмування Python. Розроблено програмні модулі, які було реалізовано у хмарному сервісі Google Colab. В результаті досліджень підтверджено заявлену у роботах інших авторів точність розпізнавання >93% Fashion MNIST DATASET та отримано покращену точність розпізнавання в 94,16% для однієї з обраних моделей. Обґрунтовано вплив параметру batch_size на точність розпізнавання, обрано значення batch_size відповідно найкращому результату розпізнавання Fashion MNIST DATASET. Продемонстровано, що збільшення кількості даних для навчання покращує точність розпізнавання при використанні параметра valid_data==(X_test, X_test_ labels) замість valid_split для даних навчання. Наведені результати чисельного експеримента, які підтверджують важливість та корисність застосування методів регуляризації для вирішення проблеми перенавчання: налаштування шарів Dropout дозволило покращити точність розпізнавання.
2
Kadam, Shivam S., Amol C. Adamuthe, and Ashwini B. Patil. "CNN Model for Image Classification on MNIST and Fashion-MNIST Dataset." Journal of scientific research 64, no. 02 (2020): 374–84. http://dx.doi.org/10.37398/jsr.2020.640251.
Full text
3
Vattikonda, Navya, Anuj Kumar Gupta, Achuthananda Reddy Polu, Bhumeka Narra, and Dheeraj Varun Kumar Reddy Buddula. "Leveraging Deep Learning for Personalized Fashion Recommendations Using Fashion MNIST." International Journal of Multidisciplinary Research in Science and Business 1, no. 07 (2024): 54–77. https://doi.org/10.63665/ijmrsb.v1i07.08.
Full textAbstract:
People are particularly consciousof their clothing choices since fashion has a big influence on daily life. Large populations are usually recommended fashion goods and trends by specialists via a manual, curated process. On the other hand, e-commerce websites greatly benefit from automatic, personalized recommendation systems, which are becoming more popular. This study introduces a deep learning-based framework for personalized fashion recommendation, utilizing the Fashion-MNIST dataset as the primary data source. The dataset was dividedinto training and testing sets in a 70:30 ratio to ensure robust evaluation. CNN, Feedforward Neural Networks (FNN), and LSTM models were employed for fashion item classification. Evaluation metrics such as F1-score, recall, accuracy, precision, and loss, along with confusion matrix analysis, were utilized to assess model performance. Among the tested models, the CNN demonstrated superior performance, achieving 93.99% accuracy, with F1-score, recall, and precision all at 94% and a loss value of 0.2037. Comparative analysis further highlighted the CNN's effectiveness over FNN and LSTM models. These findings demonstrate the promise of CNN architectures for improving the precision and consistency of individualized clothing recommendation systems.
4
Vattikonda, Navya, Anuj Kumar Gupta, Achuthananda Reddy Polu, Bhumeka Narra, and Dheeraj Varun Kumar Reddy Buddula. "Leveraging Deep Learning for Personalized Fashion Recommendations Using Fashion MNIST." International Journal of Multidisciplinary Research in Science and Business 1, no. 07 (2024): 54–77. https://doi.org/10.63665/ijmrsb.v1i01.08.
Full textAbstract:
People are particularly consciousof their clothing choices since fashion has a big influence on daily life. Large populations are usually recommended fashion goods and trends by specialists via a manual, curated process. On the other hand, e-commerce websites greatly benefit from automatic, personalized recommendation systems, which are becoming more popular. This study introduces a deep learning-based framework for personalized fashion recommendation, utilizing the Fashion-MNIST dataset as the primary data source. The dataset was dividedinto training and testing sets in a 70:30 ratio to ensure robust evaluation. CNN, Feedforward Neural Networks (FNN), and LSTM models were employed for fashion item classification. Evaluation metrics such as F1-score, recall, accuracy, precision, and loss, along with confusion matrix analysis, were utilized to assess model performance. Among the tested models, the CNN demonstrated superior performance, achieving 93.99% accuracy, with F1-score, recall, and precision all at 94% and a loss value of 0.2037. Comparative analysis further highlighted the CNN's effectiveness over FNN and LSTM models. These findings demonstrate the promise of CNN architectures for improving the precision and consistency of individualized clothing recommendation systems.
5
Mukhamediev, Ravil I. "State-of-the-Art Results with the Fashion-MNIST Dataset." Mathematics 12, no. 20 (2024): 3174. http://dx.doi.org/10.3390/math12203174.
Full textAbstract:
In September 2024, the Fashion-MNIST dataset will be 7 years old. Proposed as a replacement for the well-known MNIST dataset, it continues to be used to evaluate machine learning model architectures. This paper describes new results achieved with the Fashion-MNIST dataset using classical machine learning models and a relatively simple convolutional network. We present the state-of-the-art results obtained using the CNN-3-128 convolutional network and data augmentation. The developed CNN-3-128 model containing three convolutional layers achieved an accuracy of 99.65% in the Fashion-MNIST test image set. In addition, this paper presents the results of computational experiments demonstrating the dependence between the number of adjustable parameters of the convolutional network and the maximum acceptable classification quality, which allows us to optimise the computational cost of model training.
6
Henrique, Alisson Steffens, Anita Maria da Rocha Fernandes, Rodrigo Lyra, et al. "Classifying Garments from Fashion-MNIST Dataset Through CNNs." Advances in Science, Technology and Engineering Systems Journal 6, no. 1 (2021): 989–94. http://dx.doi.org/10.25046/aj0601109.
Full text
7
Navya, Vattikonda Anuj Kumar Gupta Achuthananda Reddy Polu Bhumeka Narra and Dheeraj Varun Kumar Reddy Buddula. "Leveraging Deep Learning for Personalized Fashion Recommendations Using Fashion MNIST." International Journal of Multidisciplinary Research in Science and Business 1, no. 7 (2025): 54–77. https://doi.org/10.5281/zenodo.15349661.
Full textAbstract:
People are particularly conscious of their clothing choices since fashion has a biginfluence on daily life. Large populations are usually recommended fashion goodsand trends by specialists via a manual, curated process. On the other hand, ecommerce websites greatly benefit from automatic, personalized recommendationsystems, which are becoming more popular. This study introduces a deep learningbased framework for personalized fashion recommendation, utilizing the FashionMNIST dataset as the primary data source. The dataset was divided into trainingand testing sets in a 70:30 ratio to ensure robust evaluation. CNN, Feed forwardNeural Networks (FNN), and LSTM models were employed for fashion itemclassification. Evaluation metrics such as F1-score, recall, accuracy, precision, andloss, along with confusion matrix analysis, were utilized to assess model performance. Among the tested models, the CNN demonstrated superiorperformance, achieving 93.99% accuracy, with F1-score, recall, and precision all at94% and a loss value of 0.2037. Comparative analysis further highlighted theCNN's effectiveness over FNN and LSTM models. These findings demonstrate thepromise of CNN architectures for improving the precision and consistency ofindividualized clothing recommendation systems.
8
Navya, Vattikonda. "Leveraging Deep Learning for Personalized Fashion Recommendations Using Fashion MNIST." International Journal of Multidisciplinary Research in Science and Business 01, no. 07 (2025): 1–8. https://doi.org/10.5281/zenodo.15130896.
Full textAbstract:
—People are particularly conscious of their clothingchoices since fashion has a big influence on daily life. Largepopulations are usually recommended fashion goods and trendsby specialists via a manual, curated process. On the other hand,e-commerce websites greatly benefit from automatic,personalized recommendation systems, which are becomingmore popular. This study introduces a deep learning-basedframework for personalized fashion recommendation, utilizingthe Fashion-MNIST dataset as the primary data source. Thedataset was divided into training and testing sets in a 70:30 ratioto ensure robust evaluation. CNN, Feedforward NeuralNetworks (FNN), and LSTM models were employed for fashionitem classification. Evaluation metrics such as F1-score, recall,accuracy, precision, and loss, along with confusion matrixanalysis, were utilized to assess model performance. Among thetested models, the CNN demonstrated superior performance,achieving 93.99% accuracy, with F1-score, recall, and precisionall at 94% and a loss value of 0.2037. Comparative analysisfurther highlighted the CNN's effectiveness over FNN andLSTM models. These findings demonstrate the promise of CNNarchitectures for improving the precision and consistency ofindividualized clothing recommendation systems
9
Sumera, Sumera, R. Sirisha, Nadia Anjum, and K. Vaidehi. "Implementation of CNN and ANN for Fashion-MNIST-Dataset using Different Optimizers." Indian Journal Of Science And Technology 15, no. 47 (2022): 2639–45. http://dx.doi.org/10.17485/ijst/v15i47.1821.
Full text
10
Nocentini, Olivia, Jaeseok Kim, Muhammad Zain Bashir, and Filippo Cavallo. "Image Classification Using Multiple Convolutional Neural Networks on the Fashion-MNIST Dataset." Sensors 22, no. 23 (2022): 9544. http://dx.doi.org/10.3390/s22239544.
Full textAbstract:
As the elderly population grows, there is a need for caregivers, which may become unsustainable for society. In this situation, the demand for automated help increases. One of the solutions is service robotics, in which robots have automation and show significant promise in working with people. In particular, household settings and aged people’s homes will need these robots to perform daily activities. Clothing manipulation is a daily activity and represents a challenging area for a robot. The detection and classification are key points for the manipulation of clothes. For this reason, in this paper, we proposed to study fashion image classification with four different neural network models to improve apparel image classification accuracy on the Fashion-MNIST dataset. The network models are tested with the highest accuracy with a Fashion-Product dataset and a customized dataset. The results show that one of our models, the Multiple Convolutional Neural Network including 15 convolutional layers (MCNN15), boosted the state of art accuracy, and it obtained a classification accuracy of 94.04% on the Fashion-MNIST dataset with respect to the literature. Moreover, MCNN15, with the Fashion-Product dataset and the household dataset, obtained 60% and 40% accuracy, respectively.
11
Salemdeeb, Mohammed, and Sarp Ertürk. "Full depth CNN classifier for handwritten and license plate characters recognition." PeerJ Computer Science 7 (June 18, 2021): e576. http://dx.doi.org/10.7717/peerj-cs.576.
Full textAbstract:
Character recognition is an important research field of interest for many applications. In recent years, deep learning has made breakthroughs in image classification, especially for character recognition. However, convolutional neural networks (CNN) still deliver state-of-the-art results in this area. Motivated by the success of CNNs, this paper proposes a simple novel full depth stacked CNN architecture for Latin and Arabic handwritten alphanumeric characters that is also utilized for license plate (LP) characters recognition. The proposed architecture is constructed by four convolutional layers, two max-pooling layers, and one fully connected layer. This architecture is low-complex, fast, reliable and achieves very promising classification accuracy that may move the field forward in terms of low complexity, high accuracy and full feature extraction. The proposed approach is tested on four benchmarks for handwritten character datasets, Fashion-MNIST dataset, public LP character datasets and a newly introduced real LP isolated character dataset. The proposed approach tests report an error of only 0.28% for MNIST, 0.34% for MAHDB, 1.45% for AHCD, 3.81% for AIA9K, 5.00% for Fashion-MNIST, 0.26% for Saudi license plate character and 0.97% for Latin license plate characters datasets. The license plate characters include license plates from Turkey (TR), Europe (EU), USA, United Arab Emirates (UAE) and Kingdom of Saudi Arabia (KSA).
12
Sumera, Sirisha R, Anjum Nadia, and Vaidehi K. "Implementation of CNN and ANN for Fashion-MNIST-Dataset using Different Optimizers." Indian Journal of Science and Technology 15, no. 47 (2022): 2639–45. https://doi.org/10.17485/IJST/v15i47.1821.
Full textAbstract:
Abstract <strong>Objectives:</strong> The paper presents application of convolution neural network and artificial neural network for image classification problem for clothing dataset along with their performance comparison against different optimizers. The major objective of this paper is to perform image classification on fashion-mnist clothing dataset images. <strong>Methods:</strong> The methods used here are, the traditional ANN and CNN. Here image classification is performed on Fashion-mnist, clothing dataset using CNN and ANN with different optimizers. The performance of the working of ANN and CNN in classifying images from fashion-mnist dataset is compared against different optimizers namely stochastic gradient Descent, Adagrad, RMS prop and Adam optimizer.<strong> Findings:</strong> The study found that CNN worked better than ANN yielding training accuracy of 95%, 93% and testing accuracy of 91%, 89% when used with Adam and RmsProp respectively. <strong>Novelty:</strong> The novelty of this work is to present a comparative study of image classification using CNN, ANN using different optimizers, since not many studies or research articles showed the performance comparison of traditional and convolution neural networks in image classification along with different optimizers. Since the real-world scenarios of today require enormous data to be processed, CNN can fit well to diversify applications since they highly reduce the number of parameters to be trained that speeds up the training process. Moreover, to be specific on image classification problems they require the best and most prominent features to be detected and uncovered; this can be achieved using CNN since it has the concept of convolution using filters at its Core. Hence, CNN is highly recommended for such image classification applications than the traditional artificial-neural-networks because of the aforementioned reasons. <strong>Keywords:</strong> ANN (Artificial NeuralNetworks); CNN (ConvolutionNeural Networks); Optimization; SGD (Stochastic Gradient Descent); RmsProp (Root mean Square propagation); Adam (Adaptive moment estimation); AdaGrad (Adaptive Gradient)
13
Aditya, Christian Sri Kusuma, Vinna Rahmayanti Setyaning Nastiti, Qori Raditya Damayanti, and Gian Bagus Sadewa. "Implementation of Convolutional Neural Network Method in Identifying Fashion Image." JUITA : Jurnal Informatika 11, no. 2 (2023): 195. http://dx.doi.org/10.30595/juita.v11i2.17372.
Full textAbstract:
The fashion industry has changed a lot over the years, which makes it hard for people to compare different kinds of fashion. To make it easier, different styles of clothing are tried out to find the exact and precise look desired. So, we opted to employ the Convolutional Neural Network (CNN) method for fashion classification. This approach represents one of the methodologies employed to utilize computers for the purpose of recognizing and categorizing items. The goal of this research is to see how well the Convolutional Neural Network method classifies the Fashion-MNIST dataset compared to other methods, models, and classification processes used in previous research. The information in this dataset is about different types of clothes and accessories. These items are divided into 10 categories, which include ankle boots, bags, coats, dresses, pullovers, sandals, shirts, sneakers, t-shirts, and trousers. The new classification method worked better than before on the test dataset. It had an accuracy value of 95. 92%, which is higher than in previous research. This research also uses a method called image data generator to make the Fashion MNIST image better. This method helps prevent too much focus on certain details and makes the results more accurate.
14
Chen, Zhengliang, Mingrui Lv, Shiqi Tian, and Shangyuan Yin. "Fashion-MNIST Classification Based on CNN Image Recognition." Highlights in Science, Engineering and Technology 34 (February 28, 2023): 196–202. http://dx.doi.org/10.54097/hset.v34i.5463.
Full textAbstract:
Image classification is an important part of Artificial Intelligence (AI) and involves several tasks such as image normalization, image segmentation, feature extraction etc. Convolutional Neural Network (CNN) has been proved to be an effective network in image classification. According to this study, we had used a relatively small dataset named Fashion-MNIST (i.e. 70,000 images, 10 categories) to find out the model that can have a higher accuracy with limited training samples. We have trained three classic CNN models which are DenseNet121, MobileNet, ResNet 50 respectively. After that, we comprehensively measured the performance by several norms that comes from these three different models. Finally, we had a conclusion on which could be the most efficient one in this scenario based on the test results, in order to discovery which models are the most powerful one, and worth training. Human always deal with many types of images, so people need a powerful AI to help them to recognize and categorize these images. After this study, the DenseNet121 is the most powerful model in these three - DenseNet121, MobileNet, ResNet 50, the method to determine this result is that in the whole study we used a method called control variate method, we use the same amount of images, same amount of training times, then compared the final output of these three models, in the end we discovered that the DenseNet121 is the most powerful one.
15
Rizwan, Muhammad, Hamza Naveed, and Rana Muhammad Zulqarnain. "Comparative Analysis of Neural Network Architectures for Image Classification." Journal of Intelligent Decision Making and Information Science 2 (May 14, 2025): 456–71. https://doi.org/10.59543/jidmis.v2i.11481.
Full textAbstract:
An important challenge in the growing domains of computer vision and machine learning is the precise categorisation of images, which drives the growth of innovative methodologies and methods. The proposed study will examine three distinct picture classification frameworks. The structures include a Single Layer Network (SLN), a Multi-Layer Perceptron (MLP), and a Convolutional Neural Network (CNN) based on LeCun's technique. The study will employ the MNIST digit identification dataset, the Fashion-MNIST dataset, and the CIFAR-10 dataset. Precision, recall, and F1-score are the measures used for model evaluation. Compared to the SLN and CNN models, the MLP model exhibits superior precision and recall.
16
Blusseau, Samy, Bastien Ponchon, Santiago Velasco-Forero, Jesús Angulo, and Isabelle Bloch. "Approximating morphological operators with part-based representations learned by asymmetric auto-encoders." Mathematical Morphology - Theory and Applications 4, no. 1 (2020): 64–86. http://dx.doi.org/10.1515/mathm-2020-0102.
Full textAbstract:
AbstractThis paper addresses the issue of building a part-based representation of a dataset of images. More precisely, we look for a non-negative, sparse decomposition of the images on a reduced set of atoms, in order to unveil a morphological and explainable structure of the data. Additionally, we want this decomposition to be computed online for any new sample that is not part of the initial dataset. Therefore, our solution relies on a sparse, non-negative auto-encoder, where the encoder is deep (for accuracy) and the decoder shallow (for explainability). This method compares favorably to the state-of-the-art online methods on two benchmark datasets (MNIST and Fashion MNIST) and on a hyperspectral image, according to classical evaluation measures and to a new one we introduce, based on the equivariance of the representation to morphological operators.
17
Sandag, Green Arther, and Jacquline Waworundeng. "Identifikasi Foto Fashion Dengan Menggunakan Convolutional Neural Network (CNN)." CogITo Smart Journal 7, no. 2 (2021): 305–14. http://dx.doi.org/10.31154/cogito.v7i2.340.305-314.
Full textAbstract:
Perkembangan teknologi sekarang ini berdampak pada banyak hal, salah satunya ialah pada bidang fashion. Penggunaan Artificial Intelligence dan juga deep learning dapat dimanfaatkan dalam bidang fashion, salah satu contohnya adalah pengenalan objek clothing. Pada penelitian ini, peneliti mengidentifikasi mode pakaian dengan menggunakan metode Convolutional Neural Network (CNN), dan library Tensorflow, serta menggunakan Fashion MNIST dataset untuk menguji kemampuan CNN model. Hasil yang didapatkan saat pengujian dengan menggunakan berbagai convolutional layer sekuensial yang kompleks, didapati dua hasil yang sedikit berbeda. Pengujian pada model pertama, terjadi overfitting, sehingga menghasilkan akurasi sebesar 91%. Pada pengujian kedua, dengan penambahan Dropout layers, menghasilkan akurasi yang lebih baik, yaitu sebesar 93%. Melihat dari hasil yang didapatkan, penggunaan CNN dalam mengidentifikasi mode pakaian cukup sesuai karena dapat mencapai akurasi hingga 93%. Kata kunci — Deep Learning, Pengenalan objek , Convolutional Neural Network (CNN), Tensorflow, Fashion MNIST
18
Vijayaraj, A., P. T. Vasanth Raj, R. Jebakumar, et al. "Deep Learning Image Classification for Fashion Design." Wireless Communications and Mobile Computing 2022 (June 14, 2022): 1–13. http://dx.doi.org/10.1155/2022/7549397.
Full textAbstract:
Fashion has always been an essential feature in our daily routine. It also plays a significant role in everyone’s lives. In this research, convolutional neural networks (CNN) were used to train images of different fashion styles, which were attempted to be predicted with a high success rate. Deep learning has been widely applied in a variety of fields recently. A CNN is a deep neural network that delivers the most accurate answers when tackling real-world situations. Apparel manufacturers have employed CNN to tackle various difficulties on their e-commerce sites, including clothing recognition, search, and suggestion. A set of photos from the Fashion-MNIST dataset is used to train a series of CNN-based deep learning architectures to distinguish between photographs. CNN design, batch normalization, and residual skip connections reduce the time it takes to learn. The CNN model’s findings are evaluated using the Fashion-MNIST datasets. In this paper, classification is done with a convolutional layer, filter size, and ultimately connected layers. Experiments are run with different activation functions, optimizers, learning rates, dropout rates, and batch sizes. The results showed that the choice of activation function, optimizer, and dropout rate impacts the correctness of the results.
19
Mousa, Sura Hamed, Narjis Mezaal Shati, and Nageswari Sakthivadivel. "DeepRing: Convolution Neural Network based on Blockchain Technology." Al-Mustansiriyah Journal of Science 35, no. 2 (2024): 61–69. http://dx.doi.org/10.23851/mjs.v35i2.1476.
Full textAbstract:
Background: This paper addresses specific challenges in predictive modeling, namely transparency issues, susceptibility to data manipulation, and fairness concerns. To overcome these obstacles, the study introduces DeepRing, approach that combines Convolutional Neural Networks (CNNs) and blockchain technology. Objective: DeepRing aims to enhance prediction integrity, data security, and fairness, thereby improving the ethical considerations, reliability, and accountability of predictive models. Methods: involves iterative training of a CNN model on five diverse datasets, including CIFAR-10, Fashion-MNIST, MNIST, CIFAR-100, and a Hands dataset. The CNN architecture incorporates Conv2D layers, MaxPooling2D layers, and Dense layers. Training metrics such as accuracy and sparse categorical cross-entropy loss are monitored, with the Adam optimizer employed. While achieving high accuracy on Plam (0.5300), MNIST (0.9978) and Fashion MNIST (0.9673), DeepRing exhibits moderate performance on CIFAR-10 (0.9296) and lower accuracy on CIFAR-100 (0.5973). Results: demonstrate the effectiveness of DeepRing in improving accuracy and enhancing model performance across various datasets. However, further development and validation are essential for successful model implementation, further development and validation are essential for successful model implementation. Conclusions: Introduces DeepRing as an innovative solution to address key challenges in predictive modeling, specifically focusing on transparency issues, susceptibility to data manipulation, and fairness concerns. By combining Convolutional Neural Networks (CNNs) with blockchain technology, DeepRing aims to elevate prediction integrity, enhance data security, and promote fairness, thereby contributing to the improvement of ethical considerations, reliability, and accountability in predictive modelling.
20
Hadiyan, Muhammad Ribhan, Firdaniza Firdaniza, and Herlina Napitupulu. "Ekstraksi Fitur Berdasarkan Fuzzy Restricted Boltzmann Machine Pada Klasifikasi Fashion-MNIST Dengan Dan Tanpa Noise." SisInfo 6, no. 2 (2024): 18–27. https://doi.org/10.37278/sisinfo.v6i2.876.
Full textAbstract:
Mixed accelerated learning method based on a Fuzzy Restricted Boltzmann Machine merupakan metode ekstraksi fitur pada gambar yang relatif baru dan belum banyak diimplementasikan. MAFRBM memiliki kelebihan dalam melakukan ekstraksi fitur pada gambar yang memiliki noise. Pada umumnya keberadaan noise pada gambar dapat mempengaruhi hasil ekstraksi fitur secara signifikan. Pada penelitian ini dilakukan ekstraksi fitur menggunakan MAFRBM pada dataset Fashion-MNIST dengan dan tanpa penambahan noise. Jenis noise yang ditambahkan pada gambar yaitu gaussian, salt & pepper, dan poisson. Hasil ekstraksi fitur MAFRBM kemudian diklasifikasikan menggunakan Support Vector Machine (SVM). Hasil klasifikasi yang diperoleh menunjukkan akurasi tertinggi sebesar 88,2%. Selain itu, perbandingan hasil akurasi dari klasifikasi fashion-MNIST dengan noise tidak berbeda jauh dengan gambar tanpa noise.
21
Tang, Yusi, Hanguo Cui, and Shuyong Liu. "Optimal Design of Deep Residual Network Based on Image Classification of Fashion-MNIST Dataset." Journal of Physics: Conference Series 1624 (October 2020): 052011. http://dx.doi.org/10.1088/1742-6596/1624/5/052011.
Full text
22
Dantas, Ítalo José de Medeiros, Marcelo Curth, and Aline Gabriel Freire. "Exploring databases for training models in machine learning in the Fashion industry." DAT Journal 9, no. 2 (2024): 157–74. http://dx.doi.org/10.29147/datjournal.v9i2.877.
Full textAbstract:
Growing interest in applying machine learning (ML) to fashion highlights the importance of using labeled data to develop models, facilitating research replication, and automating the analysis of new data, such as fashion show images available online. Despite this need, few studies, especially in Brazil, methodologically explore the intersection between fashion and AM. This research aims to provide an overview of online databases for training ML models. A systematic review identified 26 articles that use these databases, such as Fashion-MNIST and DeepFashion2. Content analysis revealed that these databases, including Polyvore and Fashion Image Dataset, have diverse applications, highlighting the transformative potential of AM in fashion and encouraging innovations in design, production, and marketing in the fashion industry.
23
Abd Alaziz, Hadeer M., Hela Elmannai, Hager Saleh, et al. "Enhancing Fashion Classification with Vision Transformer (ViT) and Developing Recommendation Fashion Systems Using DINOVA2." Electronics 12, no. 20 (2023): 4263. http://dx.doi.org/10.3390/electronics12204263.
Full textAbstract:
As e-commerce platforms grow, consumers increasingly purchase clothes online; however, they often need clarification on clothing choices. Consumers and stores interact through the clothing recommendation system. A recommendation system can help customers to find clothing that they are interested in and can improve turnover. This work has two main goals: enhancing fashion classification and developing a fashion recommendation system. The main objective of fashion classification is to apply a Vision Transformer (ViT) to enhance performance. ViT is a set of transformer blocks; each transformer block consists of two layers: a multi-head self-attention layer and a multilayer perceptron (MLP) layer. The hyperparameters of ViT are configured based on the fashion images dataset. CNN models have different layers, including multi-convolutional layers, multi-max pooling layers, multi-dropout layers, multi-fully connected layers, and batch normalization layers. Furthermore, ViT is compared with different models, i.e., deep CNN models, VGG16, DenseNet-121, Mobilenet, and ResNet50, using different evaluation methods and two fashion image datasets. The ViT model performs the best on the Fashion-MNIST dataset (accuracy = 95.25, precision = 95.20, recall = 95.25, F1-score = 95.20). ViT records the highest performance compared to other models in the fashion product dataset (accuracy = 98.53, precision = 98.42, recall = 98.53, F1-score = 98.46). A recommendation fashion system is developed using Learning Robust Visual Features without Supervision (DINOv2) and a nearest neighbor search that is built in the FAISS library to obtain the top five similarity results for specific images.
24
Wan, Guangquan, and Lan Yao. "LMFRNet: A Lightweight Convolutional Neural Network Model for Image Analysis." Electronics 13, no. 1 (2023): 129. http://dx.doi.org/10.3390/electronics13010129.
Full textAbstract:
Convolutional neural networks (CNNs) have transformed the landscape of image analysis and are widely applied across various fields. With their widespread adoption in fields like medical diagnosis and autonomous driving, CNNs have demonstrated powerful capabilities. Despite their success, existing models face challenges in deploying and operating in resource-constrained environments, limiting their practicality in real-world scenarios. We introduce LMFRNet, a lightweight CNN model. Its innovation resides in a multi-feature block design, effectively reducing both model complexity and computational load. Achieving an exceptional accuracy of 94.6% on the CIFAR-10 dataset, this model showcases remarkable performance while demonstrating parsimonious resource utilization. We further validate the performance of the model on the CIFAR-100, MNIST, and Fashion-MNIST datasets, demonstrating its robustness and generalizability across diverse datasets. Furthermore, we conducted extensive experiments to investigate the influence of critical hyperparameters. These experiments provided valuable insights for effective model training.
25
Lin, Ranxi, Benzhe Dai, Yingkai Zhao, Gang Chen, and Huaxiang Lu. "Constrain Bias Addition to Train Low-Latency Spiking Neural Networks." Brain Sciences 13, no. 2 (2023): 319. http://dx.doi.org/10.3390/brainsci13020319.
Full textAbstract:
In recent years, a third-generation neural network, namely, spiking neural network, has received plethora of attention in the broad areas of Machine learning and Artificial Intelligence. In this paper, a novel differential-based encoding method is proposed and new spike-based learning rules for backpropagation is derived by constraining the addition of bias voltage in spiking neurons. The proposed differential encoding method can effectively exploit the correlation between the data and improve the performance of the proposed model, and the new learning rule can take complete advantage of the modulation properties of bias on the spike firing threshold. We experiment with the proposed model on the environmental sound dataset RWCP and the image dataset MNIST and Fashion-MNIST, respectively, and assign various conditions to test the learning ability and robustness of the proposed model. The experimental results demonstrate that the proposed model achieves near-optimal results with a smaller time step by maintaining the highest accuracy and robustness with less training data. Among them, in MNIST dataset, compared with the original spiking neural network with the same network structure, we achieved a 0.39% accuracy improvement.
26
Ohta, Nozomu, Shin Kawai, and Hajime Nobuhara. "Capsule Network Extension Based on Metric Learning." Journal of Advanced Computational Intelligence and Intelligent Informatics 27, no. 2 (2023): 173–81. http://dx.doi.org/10.20965/jaciii.2023.p0173.
Full textAbstract:
A capsule network (CapsNet) is a deep learning model for image classification that provides robustness to changes in the poses of objects in the images. A capsule is a vector whose direction represents the presence, position, size, and pose of an object. However, with CapsNet, the distribution of capsules is concentrated in a class, and the number of capsules increases with the number of classes. In addition, learning is computationally expensive for a CapsNet. We proposed a method to increase the diversity of capsule directions and decrease the computational cost of CapsNet training by allowing a single capsule to represent multiple object classes. To determine the distance between classes, we used an additive angular margin loss called ArcFace. To validate the proposed method, the distribution of the capsules was determined using principal component analysis to validate the proposed method. In addition, using the MNIST, fashion-MNIST, EMNIST, SVHN, and CIFAR-10 datasets, as well as the corresponding affine-transformed datasets, we determined the accuracy and training time of the proposed method and original CapsNet. The accuracy of the proposed method improved by 8.91% on the CIFAR-10 dataset, and the training time reduced by more than 19% for each dataset compared with those of the original CapsNets.
27
Rifqie, Dary Mochamad, Dewi Fatmarani Surianto, Nurul Mukhlisah Abdal, Wahyu Hidayat M, and Hartini Ramli. "POST TRAINING QUANTIZATION IN LENET-5 ALGORITHM FOR EFFICIENT INFERENCE." Journal of Embedded Systems, Security and Intelligent Systems 3, no. 1 (2022): 60. http://dx.doi.org/10.26858/jessi.v3i1.34106.
Full textAbstract:
Ketika model jaringan saraf tiruan menjadi lebih baik , keinginan untuk mengimplementasikannya di dunia nyata semakin meningkat. Namun, konsumsi energi dan akurasi jaringan saraf tiruan sangat besar karena ukuran dan kompleksitasnya, sehingga sulit untuk diimplementasikan pada embedded devices. Kuantisasi jaringan saraf ini adalah sebuah teknik untuk dapat memecahkan masalah seperti mengurangi ukuran dan kompleksitas jaringan saraf tiruan dengan mengurangi ketepatan parameter dan aktivasi. Dengan jaringan yang lebih kecil, dimungkinkan untuk menjalankan jaringan saraf di lokasi yang diinginkan. Artikel ini mengkaji tentang kuantisasi yang telah berkembang dalam beberapa dekade terakhir. Dalam penelitian ini, kami mengimplementasikan kuantisasi dalam algoritma lenet-5, yang merupakan algoritma jaringan saraf convolutional pertama yang pernah ada, dan dievaluasi dalam dataset MNIST dan Fashion-MNIST.
28
Gao, Huimin, Qingtao Wu, Xuhui Zhao, Junlong Zhu, and Mingchuan Zhang. "FedADT: An Adaptive Method Based on Derivative Term for Federated Learning." Sensors 23, no. 13 (2023): 6034. http://dx.doi.org/10.3390/s23136034.
Full textAbstract:
Federated learning is served as a novel distributed training framework that enables multiple clients of the internet of things to collaboratively train a global model while the data remains local. However, the implement of federated learning faces many problems in practice, such as the large number of training for convergence due to the size of model and the lack of adaptivity by the stochastic gradient-based update at the client side. Meanwhile, it is sensitive to noise during the optimization process that can affect the performance of the final model. For these reasons, we propose Federated Adaptive learning based on Derivative Term, called FedADT in this paper, which incorporates adaptive step size and difference of gradient in the update of local model. To further reduce the influence of noise on the derivative term that is estimated by difference of gradient, we use moving average decay on the derivative term. Moreover, we analyze the convergence performance of the proposed algorithm for non-convex objective function, i.e., the convergence rate of 1/nT can be achieved by choosing appropriate hyper-parameters, where n is the number of clients and T is the number of iterations, respectively. Finally, various experiments for the image classification task are conducted by training widely used convolutional neural network on MNIST and Fashion MNIST datasets to verify the effectiveness of FedADT. In addition, the receiver operating characteristic curve is used to display the result of the proposed algorithm by predicting the categories of clothing on the Fashion MNIST dataset.
29
Duseja, Sweety. "Transfer learning-based Fashion Image Classification using Hybrid 2D-CNN and ImageNet Neural Network." International Journal for Research in Applied Science and Engineering Technology 9, no. 11 (2021): 1537–45. http://dx.doi.org/10.22214/ijraset.2021.39054.
Full textAbstract:
Abstract: Many algorithms have been developed as a result of recent advances in machine learning to handle a variety of challenges. In recent years, the most popular transfer learning method has allowed researchers and engineers to run experiments with minimal computing and time resources. To tackle the challenges of classification, product identification, product suggestion, and picture-based search, this research proposed a transfer learning strategy for Fashion image classification based on hybrid 2D-CNN pretrained by VGG-16 and AlexNet. Pre-processing, feature extraction, and classification are the three parts of the proposed system's implementation. We used the Fashion MNIST dataset, which consists of 50,000 fashion photos that have been classified. Training and validation datasets have been separated. In comparison to other conventional methodologies, the suggested transfer learning approach has higher training and validation accuracy and reduced loss. Keywords: Machine Learning, Transfer Learning, Convolutional Neural Network, Image Classification, VGG16, AlexNet, 2D CNN.
30
Li, Jianqiang, Weimin Shi, and Donghe Yang. "Clothing Image Classification with a Dragonfly Algorithm Optimised Online Sequential Extreme Learning Machine." Fibres and Textiles in Eastern Europe 29, no. 3(147) (2021): 91–96. http://dx.doi.org/10.5604/01.3001.0014.7793.
Full textAbstract:
This study proposes a solution for the issue of the low classification accuracy of clothing images. Using Fashion-MNIST as the clothing image dataset, we propose a clothing image classification technology based on an online sequential extreme learning machine (OSELM) optimised by the dragonfly algorithm (DA). First, we transform the Fashion-MNIST dataset into a data set that we extract from the corresponding grey image. Then, considering that the input weight and hidden layer bias of an OSELM are generated randomly, a DA is proposed to optimise the input weight and hidden layer bias of the OSELM to reduce the influence of random generation on the classification results. Finally, the optimised OSELM is applied to the clothing image classification. Compared to the other seven types of classification algorithms, the proposed clothing image classification model with the DA-optimised OSELM reached 93.98% accuracy when it contained 350 hidden nodes. Its performance was superior to other algorithms that were configured with the same number of hidden nodes. From a stability analysis of the box-plot, it was found that there were no outliers exhibited by the DA-OSELM model, whereas some other models had outliers or had lower stability compared to the model proposed, thereby validating the efficacy of the solution proposed.
31
Sharma, Rahul, and Amar Singh. "An Integrated Approach towards Efficient Image Classification Using Deep CNN with Transfer Learning and PCA." Advances in Technology Innovation 7, no. 2 (2022): 105–17. http://dx.doi.org/10.46604/aiti.2022.8538.
Full textAbstract:
In image processing, developing efficient, automated, and accurate techniques to classify images with varying intensity level, resolution, aspect ratio, orientation, contrast, sharpness, etc. is a challenging task. This study presents an integrated approach for image classification by employing transfer learning for feature selection and using principal component analysis (PCA) for feature reduction. The PCA algorithm is employed for reducing the dimensionality of the features extracted by the VGG16 model to obtain a handful of features for speeding up image reorganization. For multilayer perceptron classifiers, support vector machine (SVM) and random forest (RF) algorithms are used. The performance of the proposed approach is compared with other classifiers. The experimental results establish the supremacy of the VGG16-PCA-Multilayer perceptron model integrated approach and achieve a reorganization accuracy of 91.145%, 95.0%, 92.33%, and 98.59% on Fashion-MNIST dataset, ORL dataset of faces, corn leaf disease dataset, and rice leaf disease datasets, respectively.
32
Obaid, Mustafa Amer, and Wesam M. Jasim. "Pre-convoluted neural networks for fashion classification." Bulletin of Electrical Engineering and Informatics 10, no. 2 (2021): 750–58. http://dx.doi.org/10.11591/eei.v10i2.2750.
Full textAbstract:
In this work, concept of the fashion-MNIST images classification constructed on convolutional neural networks is discussed. Whereas, 28×28 grayscale images of 70,000 fashion products from 10 classes, with 7,000 images per category, are in the fashion-MNIST dataset. There are 60,000 images in the training set and 10,000 images in the evaluation set. The data has been initially pre-processed for resizing and reducing the noise. Then, this data is normalized for ensuring that all the data are on the same scale and this usually improves the performance. After normalizing the data, it is augmented where one image will be in three forms of output. The first output image is obtained by rotating the actual one; the second output image is obtained as acute angle image; and the third is obtained as tilt image. The new data set is of 180,000 images for training phase and 30,000 images for the testing phase. Finally, data is sent to training process as input for training model of the pre-convolution network. The pre-convolution neural network with the five layered convoluted deep neural network and do the training with the augmented data, The performance of the proposed system shows 94% accuracy where it was 93% in VGG16 and 92% in AlexNetnetworks.
33
Mustafa, Amer Obaid, and M. Jasim Wesam. "Pre-convoluted neural networks for fashion classification." Bulletin of Electrical Engineering and Informatics 10, no. 2 (2021): 750~758. https://doi.org/10.11591/eei.v10i2.2750.
Full textAbstract:
In this work, concept of the fashion-MNIST images classification constructed on convolutional neural networks is discussed. Whereas, 28×28 grayscale images of 70,000 fashion products from 10 classes, with 7,000 images per category, are in the fashion-MNIST dataset. There are 60,000 images in the training set and 10,000 images in the evaluation set. The data has been initially pre-processed for resizing and reducing the noise. Then, this data is normalized for ensuring that all the data are on the same scale and this usually improves the performance. After normalizing the data, it is augmented where one image will be in three forms of output. The first output image is obtained by rotating the actual one; the second output image is obtained as acute angle image; and the third is obtained as tilt image. The new data set is of 180,000 images for training phase and 30,000 images for the testing phase. Finally, data is sent to training process as input for training model of the pre-convolution network. The pre-convolution neural network with the five layered convoluted deep neural network and do the training with the augmented data, The performance of the proposed system shows 94% accuracy where it was 93% in VGG16 and 92% in AlexNetnetworks.
34
Ye, Fan, Xin-Gui Tang, Jia-Ying Chen, et al. "Neurosynaptic-like behavior of Ce-doped BaTiO3 ferroelectric thin film diodes for visual recognition applications." Applied Physics Letters 121, no. 17 (2022): 171901. http://dx.doi.org/10.1063/5.0120159.
Full textAbstract:
Brain-like neuromorphic computing networks based on the human brain information processing model are gradually breaking down the memory barriers caused by traditional computing frameworks. The brain-like neural system consists of electronic synapses and neurons. The multiple ferroelectric polarization switching modulated by the external electric field is well suited to simulate artificial neural synaptic weights. Therefore, ferroelectric diodes' (FDs) synapses have great advantages in building highly reliable and energy-efficient artificial neural networks. In this paper, we demonstrate the FDs synapse, which is based on rare-earth metal-doped BaTiO3 ferroelectric dielectric layer materials. This performs short-term and long-term synaptic plasticity behaviors by modulating synaptic weights using pulsed stimuli to polarize or flip ferroelectric films. In addition, convolutional neural networks were constructed on the MNIST dataset and the Fashion-MNIST dataset to check the feasibility of the device in simulating bio-visual recognition. The results expand the application of FDs' devices in the intersection of artificial intelligence and bioscience.
35
Mujeeb Ur Rehman and Zenab Bibi. "Performance Analysis of SpectralNet Algorithm on Fashion MNIST and KMNIST Datasets: A Study on Clustering Efficiency and Resource Utilization." Pakistan Journal of Engineering and Technology 8, no. 1 (2025): 47–54. https://doi.org/10.51846/vol8iss1pp47-54.
Full textAbstract:
Unsupervised learning faces an essential data clustering challenge in high dimensions where SpectralNet stands as an effective approach which combines deep learning with spectral clustering methods. A performance evaluation of SpectralNet measures its results on Fashion MNIST and KMNIST with accuracy as well as computational costs and resource demands under multiple hyperparameter configurations. The investigation examines how generalization changes with various distribution scenarios through the assessment of balanced versus unbalanced dataset splits. Higher embedding dimension values lead to superior clustering precision, whereas it demands increased processor capacity. The research reveals how SpectralNet handles accuracy-efficiency relationships in dataset analysis to demonstrate its practical capabilities for complex information.
36
Madhu, Golla, Sandeep Kautish, Khalid Abdulaziz Alnowibet, Hossam M. Zawbaa, and Ali Wagdy Mohamed. "NIPUNA: A Novel Optimizer Activation Function for Deep Neural Networks." Axioms 12, no. 3 (2023): 246. http://dx.doi.org/10.3390/axioms12030246.
Full textAbstract:
In recent years, various deep neural networks with different learning paradigms have been widely employed in various applications, including medical diagnosis, image analysis, self-driving vehicles and others. The activation functions employed in deep neural networks have a huge impact on the training model and the reliability of the model. The Rectified Linear Unit (ReLU) has recently emerged as the most popular and extensively utilized activation function. ReLU has some flaws, such as the fact that it is only active when the units are positive during back-propagation and zero otherwise. This causes neurons to die (dying ReLU) and a shift in bias. However, unlike ReLU activation functions, Swish activation functions do not remain stable or move in a single direction. This research proposes a new activation function named NIPUNA for deep neural networks. We test this activation by training on customized convolutional neural networks (CCNN). On benchmark datasets (Fashion MNIST images of clothes, MNIST dataset of handwritten digits), the contributions are examined and compared to various activation functions. The proposed activation function can outperform traditional activation functions.
37
Jeong, Jae-Cheol, Gwang-Hyun Yu, Min-Gyu Song, et al. "Selective Layer Tuning and Performance Study of Pre-Trained Models Using Genetic Algorithm." Electronics 11, no. 19 (2022): 2985. http://dx.doi.org/10.3390/electronics11192985.
Full textAbstract:
Utilizing pre-trained models involves fully or partially using pre-trained parameters as initialization. In general, configuring a pre-trained model demands practitioners’ knowledge about problems or an exhaustive trial–error experiment according to a given task. In this paper, we propose tuning trainable layers using a genetic algorithm on a pre-trained model that is fine-tuned on single-channel image datasets for a classification task. The single-channel dataset comprises images from grayscale and preprocessed audio signals transformed into a log-Mel spectrogram. Four deep-learning models used in the experimental evaluation employed the pre-trained model with the ImageNet dataset. The proposed genetic algorithm was applied to find the highest fitness for every generation to determine the selective layer tuning of the pre-trained models. Compared to the conventional fine-tuning method and random layer search, our proposed selective layer search with a genetic algorithm achieves higher accuracy, on average, by 9.7% and 1.88% (MNIST-Fashion), 1.31% and 1.14% (UrbanSound8k), and 2.2% and 0.29% (HospitalAlarmSound), respectively. In addition, our searching method can naturally be applied to various datasets of the same task without prior knowledge about the dataset of interest.
38
Wanjiku, Raphael Ngigi, Lawrence Nderu, and Michael Kimwele. "Improved transfer learning using textural features conflation and dynamically fine-tuned layers." PeerJ Computer Science 9 (September 28, 2023): e1601. http://dx.doi.org/10.7717/peerj-cs.1601.
Full textAbstract:
Transfer learning involves using previously learnt knowledge of a model task in addressing another task. However, this process works well when the tasks are closely related. It is, therefore, important to select data points that are closely relevant to the previous task and fine-tune the suitable pre-trained model’s layers for effective transfer. This work utilises the least divergent textural features of the target datasets and pre-trained model’s layers, minimising the lost knowledge during the transfer learning process. This study extends previous works on selecting data points with good textural features and dynamically selected layers using divergence measures by combining them into one model pipeline. Five pre-trained models are used: ResNet50, DenseNet169, InceptionV3, VGG16 and MobileNetV2 on nine datasets: CIFAR-10, CIFAR-100, MNIST, Fashion-MNIST, Stanford Dogs, Caltech 256, ISIC 2016, ChestX-ray8 and MIT Indoor Scenes. Experimental results show that data points with lower textural feature divergence and layers with more positive weights give better accuracy than other data points and layers. The data points with lower divergence give an average improvement of 3.54% to 6.75%, while the layers improve by 2.42% to 13.04% for the CIFAR-100 dataset. Combining the two methods gives an extra accuracy improvement of 1.56%. This combined approach shows that data points with lower divergence from the source dataset samples can lead to a better adaptation for the target task. The results also demonstrate that selecting layers with more positive weights reduces instances of trial and error in selecting fine-tuning layers for pre-trained models.
39
Gajowniczek, Krzysztof, Yitao Liang, Tal Friedman, Tomasz Ząbkowski, and Guy Van den Broeck. "Semantic and Generalized Entropy Loss Functions for Semi-Supervised Deep Learning." Entropy 22, no. 3 (2020): 334. http://dx.doi.org/10.3390/e22030334.
Full textAbstract:
The increasing size of modern datasets combined with the difficulty of obtaining real label information (e.g., class) has made semi-supervised learning a problem of considerable practical importance in modern data analysis. Semi-supervised learning is supervised learning with additional information on the distribution of the examples or, simultaneously, an extension of unsupervised learning guided by some constraints. In this article we present a methodology that bridges between artificial neural network output vectors and logical constraints. In order to do this, we present a semantic loss function and a generalized entropy loss function (Rényi entropy) that capture how close the neural network is to satisfying the constraints on its output. Our methods are intended to be generally applicable and compatible with any feedforward neural network. Therefore, the semantic loss and generalized entropy loss are simply a regularization term that can be directly plugged into an existing loss function. We evaluate our methodology over an artificially simulated dataset and two commonly used benchmark datasets which are MNIST and Fashion-MNIST to assess the relation between the analyzed loss functions and the influence of the various input and tuning parameters on the classification accuracy. The experimental evaluation shows that both losses effectively guide the learner to achieve (near-) state-of-the-art results on semi-supervised multiclass classification.
40
Qi, Xuan, Zegang Sun, Xue Mei, and Ryad Chellali. "A Lightweight Binarized Convolutional Neural Network Model for Small Memory and Low-Cost Mobile Devices." Mobile Information Systems 2023 (April 12, 2023): 1–11. http://dx.doi.org/10.1155/2023/5870630.
Full textAbstract:
In recent years, the high cost of implementing deep neural networks due to their large model size and parameter complexity has made it a challenging problem to design lightweight models that reduce application costs. The existing binarized neural networks suffer from both the large memory occupancy and the big number of trainable params they use. We propose a lightweight binarized convolutional neural network (CBCNN) model to address the multiclass classification/identification problem. We use both binary weights and activation. We show experimentally that a model using only 0.59 M trainable params is sufficient to reach about 92.94% accuracy on the GTSRB dataset, and it has similar performances compared to other methods on MNIST and Fashion-MNIST datasets. Accordingly, most arithmetic operations with bitwise operations are simplified, thus both used memory size and memory accesses are reduced by 32 times. Moreover, the color information was removed, which also reduced drastically the training time. All these together allow our architecture to run effectively and in real time on simple CPUs (rather than GPUs). Through the results we obtained, we show that despite simplifications and color information removal, our network achieves similar performances compared to classical CNNs with lower costs in both in training and embedded deployment.
41
Galio, Efrad, Husin Alatas, and Hendradi Hardhienata. "Image Reconstruction in Multimode Optical Fibers Using Nonlocal Artificial Neural Networks." Journal of Physics: Conference Series 2973, no. 1 (2025): 012011. https://doi.org/10.1088/1742-6596/2973/1/012011.
Full textAbstract:
Abstract Image transmission on multimode optical fibers has not yet been successfully implemented due to the persistent challenge of modal dispersion effects. This paper introduces a novel approach leveraging Non-local Artificial Neural Networks (ANN) and Convolutional Neural Networks (CNN) to address and mitigate these modal dispersion effects. By employing these advanced neural network models, we reconstruct input images based on the distorted output images from multimode optical fibers, showcasing a significant improvement in image transmission quality. Our simulation utilizes well-known datasets, including MNIST, Fashion-MNIST, and Muybridge Punch, to evaluate the performance of these models. The Convolutional Neural Network model achieves average PSNR values of 1015, 1829, and 4350, and SSIM values of 0.689, 0.511, and 0.397 for each dataset, respectively. In comparison, the Non-local Artificial Neural Network model attains average PSNR values of 2305, 2015, and 4376, and SSIM values of 0.388, 0.399, and 0.255. Our results demonstrate the superior capability of the proposed neural network models in reconstructing high-quality images, paving the way for practical implementations of multimode optical fiber systems in various fields, from communications to medical applications. This innovative approach represents a significant advancement in overcoming the long-standing issue of modal dispersion in optical fibers.
42
Le, Minh-Son, Thi-Nhan Pham, Thanh-Dat Nguyen, and Ik-Joon Chang. "A Variation-Aware Binary Neural Network Framework for Process Resilient In-Memory Computations." Electronics 13, no. 19 (2024): 3847. http://dx.doi.org/10.3390/electronics13193847.
Full textAbstract:
Binary neural networks (BNNs) that use 1-bit weights and activations have garnered interest as extreme quantization provides low power dissipation. By implementing BNNs as computation-in-memory (CIM), which computes multiplication and accumulations on memory arrays in an analog fashion, namely, analog CIM, we can further improve the energy efficiency to process neural networks. However, analog CIMs are susceptible to process variation, which refers to the variability in manufacturing that causes fluctuations in the electrical properties of transistors, resulting in significant degradation in BNN accuracy. Our Monte Carlo simulations demonstrate that in an SRAM-based analog CIM implementing the VGG-9 BNN model, the classification accuracy on the CIFAR-10 image dataset is degraded to below 50% under process variations in a 28 nm FD-SOI technology. To overcome this problem, we present a variation-aware BNN framework. The proposed framework is developed for SRAM-based BNN CIMs since SRAM is most widely used as on-chip memory; however, it is easily extensible to BNN CIMs based on other memories. Our extensive experimental results demonstrate that under process variation of 28 nm FD-SOI, with an SRAM array size of 128×128, our framework significantly enhances classification accuracies on both the MNIST hand-written digit dataset and the CIFAR-10 image dataset. Specifically, for the CONVNET BNN model on MNIST, accuracy improves from 60.24% to 92.33%, while for the VGG-9 BNN model on CIFAR-10, accuracy increases from 45.23% to 78.22%.
43
Dyubele, Sithembiso, Noxolo Pretty Cele, Lubabalo Mbangata, and Phirime Monyeki. "Evaluation and Comparison of Machine Learning Algorithms for Effective Image Classification with Fault-Tolerance." Advances in Artificial Intelligence and Machine Learning 04, no. 04 (2024): 3006–58. https://doi.org/10.54364/aaiml.2024.44174.
Full textAbstract:
Image classification is critical in computer vision, with numerous applications ranging from e-commerce to medical imaging. This study provides a comprehensive evaluation of traditional machine learning algorithms for image classification, implementing and analysing novel fault tolerance mechanisms amongst these algorithms. The authors compared the performance of K-Nearest Neighbors (KNN), Decision Trees, Random Forest, and XGBoost on both Fashion MNIST and CIFAR-10 datasets. The comparison was extended to include Support Vector Machine (SVM), Logistic Regression, and Naive Bayes classifiers in order to expand the evaluation of these models on the indicated datasets. Key findings demonstrated the superiority of ensemble methods, particularly XGBoost, which achieved 89.31% of accuracy on Fashion MNIST and 54.93% on CIFAR-10, consistently outperforming other models across various configurations. Random Forest exhibited robust performance as the secondbest model, reaching 87.42% and 51.64% of accuracy on the respective datasets. The significant performance gap between datasets demonstrated the challenges that traditional machine learning models face with complex image data. Implementing the fault tolerance framework in this study has also shown a remarkable effectiveness, achieved a 94.6% recovery rate while maintaining model accuracy within 0.1% of standard implementations. This was achieved with minimal computational overhead (2.3% of training time and 1.8% of memory usage), making it highly practical for production deployments. The system significantly reduced operational failures, decreasing crashes from 5.2 to 0.3 per day and increasing average uptime from 4.3 to 12.0 hours. The study also reveals important insights regarding model scalability and resource requirements, with memory usage varying significantly across models (325MB to 8,923MB). These findings provide valuable guidance for practitioners in selecting and implementing machine learning models for image classification tasks, particularly in scenarios where both performance and system reliability are critical. This research contributes to the field by demonstrating the feasibility of implementing robust fault tolerance in machine learning systems without compromising accuracy while also providing comprehensive performance comparisons across different model architectures and dataset complexities. The developed framework serves as a foundation for building more reliable machine-learning systems for real-world applications.
44
Xu, Zhipeng, Yao Ni, Mingxin Sun, Yiming Yuan, Ning Wu, and Wentao Xu. "Graphene/F16CuPc synaptic transistor for the emulation of multiplexed neurotransmission." Journal of Semiconductors 46, no. 1 (2025): 012603. https://doi.org/10.1088/1674-4926/24080035.
Full textAbstract:
Abstract We demonstrate a bipolar graphene/F16CuPc synaptic transistor (GFST) with matched p-type and n-type bipolar properties, which emulates multiplexed neurotransmission of the release of two excitatory neurotransmitters in graphene and F16CuPc channels, separately. This process facilitates fast-switching plasticity by altering charge carriers in the separated channels. The complementary neural network for image recognition of Fashion-MNIST dataset was constructed using the matched relative amplitude and plasticity properties of the GFST dominated by holes or electrons to improve the weight regulation and recognition accuracy, achieving a pattern recognition accuracy of 83.23%. These results provide new insights to the construction of future neuromorphic systems.
45
Li, Sijia. "The study for optimization strategies on the performance of DCGAN." Journal of Physics: Conference Series 2634, no. 1 (2023): 012032. http://dx.doi.org/10.1088/1742-6596/2634/1/012032.
Full textAbstract:
Abstract Since Deep Convolutional Generative Adversarial Networks (DCGAN) was proposed, it has been perceived as a model with difficulty in training due to several factors. To solve this problem, dozens of optimization strategies were presented, but none of them was compared with the others. In this paper, the author chose three representative methods, namely one-label smoothing, the two Time-Scale Update Rule (TTUR), and the Earth-Mover Distance (EMD) or Wasserstein-1 to make a comparison of the optimization effect on the DCGAN model. To be specific, these three approaches were adopted respectively while using MNIST and Fashion-MNIST as datasets. One-side label smoothing was designed to prevent overconfidence in the model by adding a penalty term in the discriminator. TTUR was a simpler update strategy that could help the model find the stationary local Nash equilibrium under mild assumptions. EMD was an alternative loss function that enabled the model to distinguish the difference while the real distribution and generated distribution were not overlapped. Contrast experiments were conducted both vertically and horizontally. The author applied these three methods with the same dataset and the same method with different datasets in order to compare the time of the model collapse, the trend of loss in line graphs, and the impact of different datasets on results. Experimental results indicated that both one-label smoothing and TTUR postponed the model collapse while EMD completely get rid of it. Furthermore, generated images may lose texture information when using more complicated datasets.
46
Su, Jun, Wei He, Yingguan Wang, Zhiyong Bu, and Tiantian Zhang. "Structure-Based Training: A Training Method Aimed at Pixel Errors for a Correlation-Coefficient-Based Neural Network." Sensors 24, no. 20 (2024): 6761. http://dx.doi.org/10.3390/s24206761.
Full textAbstract:
In research on building a one-shot learning neural network without pre-training using mass data, the limitation on the information obtained from a single training sample downgrades the performance of the network. In order to improve performance and take full advantage of the support set, in this study, we design three kinds of shadow nodes and propose a structure-based training method for a correlation-coefficient-based neural network. This training strategy focuses on branches that are not activated or inactivated as expected. In contrast to existing networks that optimize the parameters using back-propagation, the training method proposed in this paper optimizes the structure of the correlation-coefficient-based network by correcting its pixel errors. For the shadow nodes and training process based on this strategy, the intersection over union (IOU) of a detected target increases by 4.83% in the experiments when using the Fashion-Mnist dataset, increases by 4.02% when using the Omniglot dataset, and increases by 3.89% when using the Cifar-10 dataset. The samples in category “7” wrongly classified as “1” decreased by 27.32% when using the Mnist dataset after training. This training strategy, along with shadow nodes, makes the correlation-coefficient-based network a more practical model and enables the network to develop during the accumulation of reliable samples, thus making it more suitable for simple target detection projects that collect samples over time. Moreover, the shadow nodes and training method proposed in this paper supplement the non-gradient-based parameter-gaining strategy. Additionally, it is a new attempt to explore the imitation of a human’s ability to learn a new pattern from a low number of references.
47
Swain, Debabrata, Kaxit Pandya, Jay Sanghvi, and Yugandhar Manchala. "An Intelligent Fashion Object Classification Using CNN." EAI Endorsed Transactions on Industrial Networks and Intelligent Systems 10, no. 4 (2023): e2. http://dx.doi.org/10.4108/eetinis.v10i4.4315.
Full textAbstract:
Every year the count of visually impaired people is increasing drastically around the world. At present time, approximately 2.2 billion people are suffering from visual impairment. One of the major areas where our model will affect public life is the area of house assistance for specially-abled persons. Because of visual improvement, these people face lots of issues. Hence for this group of people, there is a high need for an assistance system in terms of object recognition. For specially-abled people sometimes it becomes really difficult to identify clothing-related items from one another because of high similarity. For better object classification we use a model which includes computer vision and CNN. Computer vision is the area of AI that helps to identify visual objects. Here a CNN-based model is used for better classification of clothing and fashion items. Another model known as Lenet is used which has a stronger architectural structure. Lenet is a multi-layer convolution neural network that is mainly used for image classification tasks. For model building and validation MNIST fashion dataset is used.
48
Parmar, Vivek, Sandeep Kaur Kingra, Shubham Negi, and Manan Suri. "Analysis of VMM computation strategies to implement BNN applications on RRAM arrays." APL Machine Learning 1, no. 2 (2023): 026108. http://dx.doi.org/10.1063/5.0139583.
Full textAbstract:
The growing interest in edge-AI solutions and advances in the field of quantized neural networks have led to hardware efficient binary neural networks (BNNs). Extreme BNNs utilize only binary weights and activations, making them more memory efficient. Such networks can be realized using exclusive-NOR (XNOR) gates and popcount circuits. The analog in-memory realization of BNNs utilizing emerging non-volatile memory devices has been widely explored recently. However, most realizations typically use 2T-2R synapses, resulting in sub-optimal area utilization. In this study, we investigate alternate computation mapping strategies to realize BNN using selectorless resistive random access memory arrays. A new differential computation scheme that shows a comparable performance with the well-established XNOR computation strategy is proposed. Through extensive experimental characterization, BNN implementation using a non-filamentary bipolar oxide-based random access memory device-based crossbar is demonstrated for two datasets: (i) experimental characterization was performed on a thermal-image based Rock-Paper-Scissors dataset to analyze the impact of sneak-paths with real-hardware experiments. (ii) Large-scale BNN simulations on the Fashion-MNIST dataset with multi-level cell characteristics of non-filamentary devices are performed to demonstrate the impact of device non-idealities.
49
Huang, Lan, Jia Zeng, Shiqi Sun, Wencong Wang, Yan Wang, and Kangping Wang. "Coarse-Grained Pruning of Neural Network Models Based on Blocky Sparse Structure." Entropy 23, no. 8 (2021): 1042. http://dx.doi.org/10.3390/e23081042.
Full textAbstract:
Deep neural networks may achieve excellent performance in many research fields. However, many deep neural network models are over-parameterized. The computation of weight matrices often consumes a lot of time, which requires plenty of computing resources. In order to solve these problems, a novel block-based division method and a special coarse-grained block pruning strategy are proposed in this paper to simplify and compress the fully connected structure, and the pruned weight matrices with a blocky structure are then stored in the format of Block Sparse Row (BSR) to accelerate the calculation of the weight matrices. First, the weight matrices are divided into square sub-blocks based on spatial aggregation. Second, a coarse-grained block pruning procedure is utilized to scale down the model parameters. Finally, the BSR storage format, which is much more friendly to block sparse matrix storage and computation, is employed to store these pruned dense weight blocks to speed up the calculation. In the following experiments on MNIST and Fashion-MNIST datasets, the trend of accuracies with different pruning granularities and different sparsity is explored in order to analyze our method. The experimental results show that our coarse-grained block pruning method can compress the network and can reduce the computational cost without greatly degrading the classification accuracy. The experiment on the CIFAR-10 dataset shows that our block pruning strategy can combine well with the convolutional networks.
50
Seong-Yoon Shin, Gwanghyun Jo, and Guangxing Wang. "A Novel Method for Fashion Clothing Image Classification Based on Deep Learning." Journal of Information and Communication Technology 22, no. 1 (2023): 127–48. http://dx.doi.org/10.32890/jict2023.22.1.6.
Full textAbstract:
Image recognition and classification is a significant research topic in computational vision and widely used computer technology. Themethods often used in image classification and recognition tasks are based on deep learning, like Convolutional Neural Networks(CNNs), LeNet, and Long Short-Term Memory networks (LSTM). Unfortunately, the classification accuracy of these methods isunsatisfactory. In recent years, using large-scale deep learning networks to achieve image recognition and classification canimprove classification accuracy, such as VGG16 and Residual Network (ResNet). However, due to the deep network hierarchyand complex parameter settings, these models take more time in the training phase, especially when the sample number is small, which can easily lead to overfitting. This paper suggested a deep learning-based image classification technique based on a CNN model and improved convolutional and pooling layers. Furthermore, the study adopted the approximate dynamic learning rate update algorithm in the model training to realize the learning rate’s self-adaptation, ensure the model’s rapid convergence, and shorten the training time. Using the proposed model, an experiment was conducted on the Fashion-MNIST dataset, taking 6,000 images as the training dataset and 1,000 images as the testing dataset. In actual experiments, the classification accuracy of the suggested method was 93 percent, 4.6 percent higher than that of the basic CNN model. Simultaneously, the study compared the influence of the batch size of model training on classification accuracy. Experimental outcomes showed this model is very generalized in fashion clothing image classification tasks.