Relevant bibliographies by topics / MXNET

Contents

  1. Journal articles
  2. Dissertations / Theses
  3. Books
  4. Conference papers

Academic literature on the topic 'MXNET'

Author: Grafiati
Published: 4 June 2021
Last updated: 18 June 2025

Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles

Select a source type:

Consult the lists of relevant articles, books, theses, conference reports, and other scholarly sources on the topic 'MXNET.'

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.

Journal articles on the topic "MXNET"

1

Li, Mingfan, Ke Wen, Han Lin, et al. "Improving the Performance of Distributed MXNet with RDMA." International Journal of Parallel Programming 47, no. 3 (2019): 467–80. http://dx.doi.org/10.1007/s10766-018-00623-w.

Full text
APA, Harvard, Vancouver, ISO, and other styles
2

Omar, Alexander Ruiz-Vivanco. "Deep Learning as a predictive model to classify handwritten digits." Latin-American Journal of Computing 4, no. 3 (2017): 73–78. https://doi.org/10.5281/zenodo.5764688.

Full text
Abstract:
In this research work, the results of applying DeepLearning prediction models to identify the digit of an image,that contains a handwritten number of the MNIST database, arepresented. This set of dataset was acquired from the competitionof Kaggle: Digit Recognizer. The following process was applied:First, image preprocessing techniques were used, which focuson obtaining a pretty clear image and to reduce the size ofthe same, these objectives that are achieved with Otsu Method,transformed from Haar Wavelet and the Principal ComponentAnalysis (PCA), thus obtaining as a result, one set of new datasetto be evaluated. Second, the Deep Learning MxNET and H2omodels, which were executed in the statistical language R, wereapplied to these datasets obtained, this way, several predictionswere acquired. Finally, the best obtained predictions in theexperiment were sent to the Digit Recognizer competition, andthe results of this evaluation scored 99,129% of prediction.
APA, Harvard, Vancouver, ISO, and other styles
3

Ramasamy Reddy, Viswanathan, Sukham Romen Singh, Elangovan Guruva Reddy, E. Punarselvam E. Punarselvam, and T. Vengatesh T. Vengatesh. "AI-Driven Enhancement of Spam Detection in SMS and Email Using AWS Leveraging Deep Spam Model." Journal of Neonatal Surgery 14, no. 15S (2025): 1458–68. https://doi.org/10.63682/jns.v14i15s.3865.

Full text
Abstract:
This study proposes a revolutionary strategy to enhance spam detection in SMS and email communications by integrating the powerful AWS cloud architecture with cutting- edge artificial intelligence (AI) approaches. The project seeks to produce a highly effective system that can discriminate between incoming messages that are spam and those that are legitimate (ham) by utilizing machine learning models that were created on the customized Amazon Sage Maker platform. The solution is deliberately developed and incorporates crucial pieces including AWS Lambda functions, Simple Email Service (SES), S3 buckets for data storage, and the trustworthy MXNet framework for model training and deployment. The suggested solution contains an extensive procedure that combines expensive pre processing, complex feature extraction approaches, hard model training processes, and seamless real- time message classification. The study's experimental findings clearly illustrate the remarkable efficacy of the offered strategy in accurately identifying and categorizing spam messages, considerably enhancing communication security and dependability overall. This research fulfills the highest conference standards, as it includes a full investigation of the topic coupled with practical application and real-world repercussions
APA, Harvard, Vancouver, ISO, and other styles
4

Elshawi, Radwa, Abdul Wahab, Ahmed Barnawi, and Sherif Sakr. "DLBench: a comprehensive experimental evaluation of deep learning frameworks." Cluster Computing 24, no. 3 (2021): 2017–38. http://dx.doi.org/10.1007/s10586-021-03240-4.

Full text
Abstract:
AbstractDeep Learning (DL) has achieved remarkable progress over the last decade on various tasks such as image recognition, speech recognition, and natural language processing. In general, three main crucial aspects fueled this progress: the increasing availability of large amount of digitized data, the increasing availability of affordable parallel and powerful computing resources (e.g., GPU) and the growing number of open source deep learning frameworks that facilitate and ease the development process of deep learning architectures. In practice, the increasing popularity of deep learning frameworks calls for benchmarking studies that can effectively evaluate and understand the performance characteristics of these systems. In this paper, we conduct an extensive experimental evaluation and analysis of six popular deep learning frameworks, namely, TensorFlow, MXNet, PyTorch, Theano, Chainer, and Keras, using three types of DL architectures Convolutional Neural Networks (CNN), Faster Region-based Convolutional Neural Networks (Faster R-CNN), and Long Short Term Memory (LSTM). Our experimental evaluation considers different aspects for its comparison including accuracy, training time, convergence and resource consumption patterns. Our experiments have been conducted on both CPU and GPU environments using different datasets. We report and analyze the performance characteristics of the studied frameworks. In addition, we report a set of insights and important lessons that we have learned from conducting our experiments.
APA, Harvard, Vancouver, ISO, and other styles
5

Cynthia, Hayat, and Aang Soenandi Iwan. "Deep learning model for detection acute cardiogenic pulmonary edema in cases of preeclampsia." IAES International Journal of Artificial Intelligence (IJ-AI) 13, no. 4 (2024): 4806–12. https://doi.org/10.11591/ijai.v13.i4.pp4806-4812.

Full text
Abstract:
The physiological changes during the pregnancy period increase the risk of developing pulmonary edema and acute respiratory failure. This condition falls under critical medical emergencies associated with maternal mortality. This study utilized a convolutional neural networks (CNN) architectural model employing chest Xray dataset images. CNN utilizes the convolution process by moving a convolutional kernel of a certain size across an image, allowing the computer to derive new representative information from the multiplication of portions of the image with the utilized filter. To simplify, the vanishing gradient issue occurs when information dissipates before reaching its destination due to the lengthy path between input and output layers. This study was developed model for detection acute cardiogenic pulmonary Edema in pre-eclampsia cases using chest Xray images, implemented using PyTorch, Keras, and MxNet. The validated model achieved its optimum with accuracy 90.65% and binary cross-entropy loss (BCELoss) value of 0.4538. It exhibited an improved sensitivity value of 83.514% using a 5% dataset and a specificity value of 57.273%. This indicates an increase in sensitivity value by 83.514% using a 5% data set and a specificity value of 57.273%. The research results demonstrate an improvement in accuracy compared to several similar studies that also utilized CNN models.
APA, Harvard, Vancouver, ISO, and other styles
6

Hayat, Cynthia, and Iwan Aang Soenandi. "Deep learning model for detection acute cardiogenic pulmonary edema in cases of preeclampsia." IAES International Journal of Artificial Intelligence (IJ-AI) 13, no. 4 (2024): 4806. http://dx.doi.org/10.11591/ijai.v13.i4.pp4806-4812.

Full text
Abstract:
<span lang="EN-US">The physiological changes during the pregnancy period increase the risk of developing pulmonary edema and acute respiratory failure. This condition falls under critical medical emergencies associated with maternal mortality. This study utilized a convolutional neural networks (CNN) architectural model employing chest Xray dataset images. CNN utilizes the convolution process by moving a convolutional kernel of a certain size across an image, allowing the computer to derive new representative information from the multiplication of portions of the image with the utilized filter.</span><span lang="EN-US">To simplify, the vanishing gradient issue occurs when information dissipates before reaching its destination due to the lengthy path between input and output layers. This study was developed model for detection acute cardiogenic pulmonary Edema in pre-eclampsia cases using chest Xray images, implemented using PyTorch, Keras, and MxNet. The validated model achieved its optimum with accuracy 90.65% and binary cross-entropy loss (BCELoss) value of 0.4538. It exhibited an improved sensitivity value of 83.514% using a 5% dataset and a specificity value of 57.273%. This indicates an increase in sensitivity value by 83.514% using a 5% data set and a specificity value of 57.273%. The research results demonstrate an improvement in accuracy compared to several similar studies that also utilized CNN models.</span>
APA, Harvard, Vancouver, ISO, and other styles
7

Jeon, Eunjoo, Kyusam Oh, Soonhwan Kwon, et al. "A Lightweight Deep Learning Model for Fast Electrocardiographic Beats Classification With a Wearable Cardiac Monitor: Development and Validation Study." JMIR Medical Informatics 8, no. 3 (2020): e17037. http://dx.doi.org/10.2196/17037.

Full text
Abstract:
Background Electrocardiographic (ECG) monitors have been widely used for diagnosing cardiac arrhythmias for decades. However, accurate analysis of ECG signals is difficult and time-consuming work because large amounts of beats need to be inspected. In order to enhance ECG beat classification, machine learning and deep learning methods have been studied. However, existing studies have limitations in model rigidity, model complexity, and inference speed. Objective To classify ECG beats effectively and efficiently, we propose a baseline model with recurrent neural networks (RNNs). Furthermore, we also propose a lightweight model with fused RNN for speeding up the prediction time on central processing units (CPUs). Methods We used 48 ECGs from the MIT-BIH (Massachusetts Institute of Technology-Beth Israel Hospital) Arrhythmia Database, and 76 ECGs were collected with S-Patch devices developed by Samsung SDS. We developed both baseline and lightweight models on the MXNet framework. We trained both models on graphics processing units and measured both models’ inference times on CPUs. Results Our models achieved overall beat classification accuracies of 99.72% for the baseline model with RNN and 99.80% for the lightweight model with fused RNN. Moreover, our lightweight model reduced the inference time on CPUs without any loss of accuracy. The inference time for the lightweight model for 24-hour ECGs was 3 minutes, which is 5 times faster than the baseline model. Conclusions Both our baseline and lightweight models achieved cardiologist-level accuracies. Furthermore, our lightweight model is competitive on CPU-based wearable hardware.
APA, Harvard, Vancouver, ISO, and other styles
8

Krichen, Moez. "Convolutional Neural Networks: A Survey." Computers 12, no. 8 (2023): 151. http://dx.doi.org/10.3390/computers12080151.

Full text
Abstract:
Artificial intelligence (AI) has become a cornerstone of modern technology, revolutionizing industries from healthcare to finance. Convolutional neural networks (CNNs) are a subset of AI that have emerged as a powerful tool for various tasks including image recognition, speech recognition, natural language processing (NLP), and even in the field of genomics, where they have been utilized to classify DNA sequences. This paper provides a comprehensive overview of CNNs and their applications in image recognition tasks. It first introduces the fundamentals of CNNs, including the layers of CNNs, convolution operation (Conv_Op), Feat_Maps, activation functions (Activ_Func), and training methods. It then discusses several popular CNN architectures such as LeNet, AlexNet, VGG, ResNet, and InceptionNet, and compares their performance. It also examines when to use CNNs, their advantages and limitations, and provides recommendations for developers and data scientists, including preprocessing the data, choosing appropriate hyperparameters (Hyper_Param), and evaluating model performance. It further explores the existing platforms and libraries for CNNs such as TensorFlow, Keras, PyTorch, Caffe, and MXNet, and compares their features and functionalities. Moreover, it estimates the cost of using CNNs and discusses potential cost-saving strategies. Finally, it reviews recent developments in CNNs, including attention mechanisms, capsule networks, transfer learning, adversarial training, quantization and compression, and enhancing the reliability and efficiency of CNNs through formal methods. The paper is concluded by summarizing the key takeaways and discussing the future directions of CNN research and development.
APA, Harvard, Vancouver, ISO, and other styles
9

Dalskov, Anders, Daniel Escudero, and Marcel Keller. "Secure Evaluation of Quantized Neural Networks." Proceedings on Privacy Enhancing Technologies 2020, no. 4 (2020): 355–75. http://dx.doi.org/10.2478/popets-2020-0077.

Full text
Abstract:
AbstractWe investigate two questions in this paper: First, we ask to what extent “MPC friendly” models are already supported by major Machine Learning frameworks such as TensorFlow or PyTorch. Prior works provide protocols that only work on fixed-point integers and specialized activation functions, two aspects that are not supported by popular Machine Learning frameworks, and the need for these specialized model representations means that it is hard, and often impossible, to use e.g., TensorFlow to design, train and test models that later have to be evaluated securely. Second, we ask to what extent the functionality for evaluating Neural Networks already exists in general-purpose MPC frameworks. These frameworks have received more scrutiny, are better documented and supported on more platforms. Furthermore, they are typically flexible in terms of the threat model they support. In contrast, most secure evaluation protocols in the literature are targeted to a specific threat model and their implementations are only a “proof-of-concept”, making it very hard for their adoption in practice. We answer both of the above questions in a positive way:We observe that the quantization techniques supported by both TensorFlow, PyTorch and MXNet can provide models in a representation that can be evaluated securely; and moreover, that this evaluation can be performed by a general purpose MPC framework. We perform extensive benchmarks to understand the exact trade-offs between different corruption models, network sizes and efficiency. These experiments provide an interesting insight into cost between active and passive security, as well as honest and dishonest majority. Our work shows then that the separating line between existing ML frameworks and existing MPC protocols may be narrower than implicitly suggested by previous works.
APA, Harvard, Vancouver, ISO, and other styles
10

Yin, Lujia, Yiming Zhang, Zhaoning Zhang, Yuxing Peng, and Peng Zhao. "ParaX." Proceedings of the VLDB Endowment 14, no. 6 (2021): 864–77. http://dx.doi.org/10.14778/3447689.3447692.

Full text
Abstract:
Despite the fact that GPUs and accelerators are more efficient in deep learning (DL), commercial clouds like Facebook and Amazon now heavily use CPUs in DL computation because there are large numbers of CPUs which would otherwise sit idle during off-peak periods. Following the trend, CPU vendors have not only released high-performance many-core CPUs but also developed efficient math kernel libraries. However, current DL platforms cannot scale well to a large number of CPU cores, making many-core CPUs inefficient in DL computation. We analyze the memory access patterns of various layers and identify the root cause of the low scalability, i.e., the per-layer barriers that are implicitly imposed by current platforms which assign one single instance (i.e., one batch of input data) to a CPU. The barriers cause severe memory bandwidth contention and CPU starvation in the access-intensive layers (like activation and BN). This paper presents a novel approach called ParaX, which boosts the performance of DL on many-core CPUs by effectively alleviating bandwidth contention and CPU starvation. Our key idea is to assign one instance to each CPU core instead of to the entire CPU, so as to remove the per-layer barriers on the executions of the many cores. ParaX designs an ultralight scheduling policy which sufficiently overlaps the access-intensive layers with the compute-intensive ones to avoid contention, and proposes a NUMA-aware gradient server mechanism for training which leverages shared memory to substantially reduce the overhead of per-iteration parameter synchronization. We have implemented ParaX on MXNet. Extensive evaluation on a two-NUMA Intel 8280 CPU shows that ParaX significantly improves the training/inference throughput for all tested models (for image recognition and natural language processing) by 1.73X ~ 2.93X.
APA, Harvard, Vancouver, ISO, and other styles
More sources

Dissertations / Theses on the topic "MXNET"

1

Yin, Jiaqi. "Measurement of machine learning performance with different condition and hyperparameter." The Ohio State University, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=osu1587693436870594.

Full text
APA, Harvard, Vancouver, ISO, and other styles
2

Sunesson, Albin. "Establishing Effective Techniques for Increasing Deep Neural Networks Inference Speed." Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-213833.

Full text
Abstract:
Recent trend in deep learning research is to build ever more deep networks (i.e. increase the number of layers) to solve real world classification/optimization problems. This introduces challenges for applications with a latency dependence. The problem arises from the amount of computations that needs to be performed for each evaluation. This is addressed by reducing inference speed. In this study we analyze two different methods for speeding up the evaluation of deep neural networks. The first method reduces the number of weights in a convolutional layer by decomposing its convolutional kernel. The second method lets samples exit a network through early exit branches when classifications are certain. Both methods were evaluated on several network architectures with consistent results. Convolutional kernel decomposition shows 20-70% speed up with no more than 1% loss in classification accuracy in setups evaluated. Early exit branches show up to 300% speed up with no loss in classification accuracy when evaluated on CPUs.<br>De senaste årens trend inom deep learning har varit att addera fler och fler lager till neurala nätverk. Det här introducerar nya utmaningar i applikationer med latensberoende. Problemet uppstår från mängden beräkningar som måste utföras vid varje evaluering. Detta adresseras med en reducering av inferenshastigheten. Jag analyserar två olika metoder för att snabba upp evalueringen av djupa neurala näverk. Den första metoden reducerar antalet vikter i ett faltningslager via en tensordekomposition på dess kärna. Den andra metoden låter samples lämna nätverket via tidiga förgreningar när en klassificering är säker. Båda metoderna utvärderas på flertalet nätverksarkitekturer med konsistenta resultat. Dekomposition på fältningskärnan visar 20-70% hastighetsökning med mindre än 1% försämring av klassifikationssäkerhet i evaluerade konfigurationer. Tidiga förgreningar visar upp till 300% hastighetsökning utan någon försämring av klassifikationssäkerhet när de evalueras på CPU.
APA, Harvard, Vancouver, ISO, and other styles

Books on the topic "MXNET"

1

Perez-Torres, Andres. Deep Learning with MXNet Cookbook: Deep Dive into a Variety of Recipes to Build, Train, and Deploy Scalable AI Models on Apache MXNet. Packt Publishing, Limited, 2023.

Find full text
APA, Harvard, Vancouver, ISO, and other styles
2

R Deep Learning Cookbook: Solve complex neural net problems with TensorFlow, H2O and MXNet. Packt Publishing - ebooks Account, 2017.

Find full text
APA, Harvard, Vancouver, ISO, and other styles
3

Wiley, Joshua F., Yuxi (Hayden) Liu, Pablo Maldonado, and Mark Hodnett. Deep Learning with R for Beginners: Design Neural Network Models in R 3. 5 Using TensorFlow, Keras, and MXNet. Packt Publishing, Limited, 2019.

Find full text
APA, Harvard, Vancouver, ISO, and other styles
4

R Deep Learning Essentials: A step-by-step guide to building deep learning models using TensorFlow, Keras, and MXNet, 2nd Edition. Packt Publishing, 2018.

Find full text
APA, Harvard, Vancouver, ISO, and other styles

Conference papers on the topic "MXNET"

1

Wang, Weiwei. "Russian Vocabulary Pronunciation System based on Computerized MXNet." In 2024 7th International Conference on Education, Network and Information Technology (ICENIT). IEEE, 2024. https://doi.org/10.1109/icenit61951.2024.00029.

Full text
APA, Harvard, Vancouver, ISO, and other styles
2

Kim, Seongsoo, Hayden Wimmer, and Jongyeop Kim. "Analysis of Deep Learning Libraries: Keras, PyTorch, and MXnet." In 2022 IEEE/ACIS 20th International Conference on Software Engineering Research, Management and Applications (SERA). IEEE, 2022. http://dx.doi.org/10.1109/sera54885.2022.9806734.

Full text
APA, Harvard, Vancouver, ISO, and other styles
3

Lv, Baocai, Bing Liu, Fang Liu, Nong Xiao, and Zhiguang Chen. "RM-KVStore: New MXNet KVStore to Accelerate Transfer Performancewith RDMA." In 2018 IEEE Symposium on Computers and Communications (ISCC). IEEE, 2018. http://dx.doi.org/10.1109/iscc.2018.8538485.

Full text
APA, Harvard, Vancouver, ISO, and other styles
4

Jain, Arpan, Ammar Ahmad Awan, Hari Subramoni, and Dhabaleswar K. Panda. "Scaling TensorFlow, PyTorch, and MXNet using MVAPICH2 for High-Performance Deep Learning on Frontera." In 2019 IEEE/ACM Third Workshop on Deep Learning on Supercomputers (DLS). IEEE, 2019. http://dx.doi.org/10.1109/dls49591.2019.00015.

Full text
APA, Harvard, Vancouver, ISO, and other styles
5

Gong, Linyuan, Jiayi Wang, and Alvin Cheung. "ADELT: Transpilation between Deep Learning Frameworks." In Thirty-Third International Joint Conference on Artificial Intelligence {IJCAI-24}. International Joint Conferences on Artificial Intelligence Organization, 2024. http://dx.doi.org/10.24963/ijcai.2024/694.

Full text
Abstract:
We propose the Adversarial DEep Learning Transpiler (ADELT), a novel approach to source-to-source transpilation between deep learning frameworks. ADELT uniquely decouples code skeleton transpilation and API keyword mapping. For code skeleton transpilation, it uses few-shot prompting on large language models (LLMs), while for API keyword mapping, it uses contextual embeddings from a code-specific BERT. These embeddings are trained in a domain-adversarial setup to generate a keyword translation dictionary. ADELT is trained on an unlabeled web-crawled deep learning corpus, without relying on any hand-crafted rules or parallel data. It outperforms state-of-the-art transpilers, improving pass@1 rate by 16.2 pts and 15.0 pts for PyTorch-Keras and PyTorch-MXNet transpilation pairs respectively. We provide open access to our code at https://github.com/gonglinyuan/adelt
APA, Harvard, Vancouver, ISO, and other styles
6

Li, Shiyang, and Cheng Li. "Research on Surface Defect Detection in Submersible Pump Impellers Based on the Novel MXNet Deep Learning Model." In 2023 IEEE 3rd International Conference on Data Science and Computer Application (ICDSCA). IEEE, 2023. http://dx.doi.org/10.1109/icdsca59871.2023.10393601.

Full text
APA, Harvard, Vancouver, ISO, and other styles
7

Alecu, Alin adrian, Bujor ionel Pavaloiu, and Ciprian gabriel Dobretrifan. "A NOVEL FREE CLOUD SERVICE FOR MACHINE LEARNING AND BEYOND." In eLSE 2019. Carol I National Defence University Publishing House, 2019. http://dx.doi.org/10.12753/2066-026x-19-030.

Full text
Abstract:
The concept of eLearning has grown exponentially in recent years. Indeed, what was but a few decades ago exclusively reserved for classroom attendees of university institutions and their closed intranets has today become a publicly accessible source of information, with public and private-owned eLearning MOOC (Massive Open Online Courses) platforms dominating the digital landscape and bringing with them a plethora of services such as video courses, certifications and employee recruitment. Nonetheless, while adding significant public value, such platforms still fail to completely bridge the gap between academia and industry in the hands-on knowledge application sense. This is especially true for domains such as machine learning that are known to require tremendous computational resources if one wishes to transition from simplistic lab exercises to complex state-of-the-art models, which is something a MOOC - or most academic institutions for that matter - simply cannot afford. Private cloud providers have taken up the challenge here and opened their offerings to the research community with programs such as TensorFlow Research Cloud, CNTK on Microsoft Azure, MXNet on Amazon AWS, to name but a few. Nonetheless, access to such resources remains limited: either the resources themselves are scarce and consequently freely available only to a restricted research community, or they are free for a limited time, or they follow a freemium model. This paper describes a novel distributed publicly-hosted overlay network that provides free collaborative eLearning material and computational resources to its users. Unlike most of its P2P counterparts in existence today that address alternative public needs (i.e. anonymity, file sharing, ...), it offers unlimited runtime access to all the computational resources that collectively define its topology. In this respect, an example use case is one wherein a distributed machine learning system can be orchestrated by any user and freely deployed for execution on any number of available nodes within the network, without any limitation other than that of the underlining hardware and network latency performance capabilities themselves. Additionally, source code and accompanying eLearning courses are regularly published to the network and made freely available by users belonging to the academic and subject matter expert practitioner community, implementing solutions ranging from simple classroom exercises to entire production-ready systems. The network is collaborative by design, such that learning material can be collectively worked on and published, training data can be shared and successfully deployed software solutions can be load-balanced and made accessible to the entire community. Moreover, the distributed network follows a decentralized model governed by a handful of master nodes that, similarly to all its other nodes, are hosted by the users themselves, and in this sense constitutes a novel cloud offering maintained by the general public rather than any given private cloud provider. Finally, we argue that beyond its educational and computational power offerings, the described network likely constitutes the best test platform for any designed software system: indeed, given the volatile nature of the network (wherein user-hosted resources can join or go offline at any moment), we claim that any software system that can "survive" and maintain its levels of SLA (service level agreement) on this network is guaranteed to perform similarly or better on a more traditional pay-per-usage cloud provider infrastructure.
APA, Harvard, Vancouver, ISO, and other styles
You might also be interested in the extended bibliographies on the topic 'MXNET' for particular source types:
Journal articles
We offer discounts on all premium plans for authors whose works are included in thematic literature selections. Contact us to get a unique promo code!

To the bibliography