The System.AI Project: Fully Managed Cross-Platform Machine Learning and Data Analysis Stack for .NET Ecosystem
- Autores: Brykin G.S.1
-
Afiliações:
- Bauman Moscow State Technical University
- Edição: Volume 73, Nº 1 (2023)
- Páginas: 64-72
- Seção: Data Mining
- URL: https://journal-vniispk.ru/2079-0279/article/view/286867
- DOI: https://doi.org/10.14357/20790279230108
- ID: 286867
Citar
Texto integral
Resumo
In recent years, machine learning technologies have become increasingly popular in widespread tasks such as image stylization, black-and-white image coloring, super-resolution of images, fake data searching, voice and image recognition. In this regard, there is a need to implement a set of tools for integrating artificial intelligence systems into applications for mobile devices, smart home devices, and home PCs. The paper describes a solution that allows developers to integrate data analysis and machine learning systems directly into a user application, which will allow to produce a lightweight, portable, and cross-platform monolithic application, which is often not possible with existing solutions. The main features of the proposed solution are the focus on the Microsoft .NET [1] ecosystem and the use of exclusively standard features of BCL and C# programming language. The implemented package of tools is completely cross-platform and hardware independent. The API is similar in many ways to its Python counterparts, which allows to quickly migrate Python codes into a .NET project.
Palavras-chave
Sobre autores
Gleb Brykin
Bauman Moscow State Technical University
Autor responsável pela correspondência
Email: glebbrykin@colorfulsoft.ru
Rússia, ul. Baumanskaya 2-ya, 5/1, Moscow, 105005
Bibliografia
- .NET homepage. Available at: https://dotnet. microsoft.com (accessed November 22, 2022)
- Mono homepage. Available at: https://www.monoproject.com (accessed November 22, 2022)
- Xamarin homepage. Available at: https://dotnet. microsoft.com/apps/xamarin (accessed November 22, 2022)
- Imageio homepage. Available at: https://github.com/ imageio/imageio (accessed November 22, 2022)
- Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, Alban Desmaison, Andreas Köpf, Edward Yang, Zach DeVito, Martin Raison, Alykhan Tejani, Sasank Chilamkurthy, Benoit Steiner, Lu Fang, Junjie Bai, Soumith Chintala. 2019. PyTorch: An Imperative Style, High-Performance Deep Learning Library. arXiv:1912.01703. Available at: https://arxiv.org/abs/1912.01703 (accessed November 22, 2022)
- convent.js homepage. Available at: https://cs.stanford.edu/people/ karpathy/convnetjs (accessed November 22, 2022)
- Martín Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dan Mane, Rajat Monga, Sherry Moore, Derek Murray, Chris Olah, Mike Schuster, Jonathon Shlens, Benoit Steiner, Ilya Sutskever, Kunal Talwar, Paul Tucker, Vincent Vanhoucke, Vijay Vasudevan, Fernanda Viegas, Oriol Vinyals, Pete Warden, Martin Wattenberg, Martin Wicke, Yuan Yu, Xiaoqiang Zheng. 2016. TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems. arXiv:1603.04467. Available at: https://arxiv.org/abs/1603.04467 (accessed November 22, 2022)
- ConvNetCS GitHub repository. Available at: https://github.com/mashmawy/ConvNetCS (accessed November 22, 2022)
- Daniel Smilkov, Nikhil Thorat, Yannick Assogba, Ann Yuan, Nick Kreeger, Ping Yu, Kangyi Zhang, Shanqing Cai, Eric Nielsen, David Soergel, Stan Bileschi, Michael Terry, Charles Nicholson, Sandeep N. Gupta, Sarah Sirajuddin, D. Sculley, Rajat Monga, Greg Corrado, Fernanda B. Viégas, Martin Wattenberg. 2019. TensorFlow. js: Machine Learning for the Web and Beyond. arXiv:1901.05350. Available at: https://arxiv.org/ abs/1901.05350 (accessed November 22, 2022)
- AlbiruniML GitHub repository. Available at: https://github.com/mashmawy/AlbiruniML (accessed November 22, 2022)
- SciSharp STACK homepage. Available at: https:// scisharp.github.io/SciSharp/ (accessed November 22, 2022)
- Demos GitHub repository. Available at: https:// github.com/ColorfulSoft/StyleTransferColorization-SuperResolution (accessed November 22, 2022)
- David W. Brown. 2018. MyCaffe: A Complete C# Re-Write of Caffe with Reinforcement Learning. arXiv:1810.02272. Available at: https://arxiv.org/ abs/1810.02272 (accessed November 22, 2022)
- Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick, S. Guadarrama and T. Darrell. 2014. Caffe: Convolution Architecture for Fast Feature Embedding. arXiv:1408.5093. Available at: https://arxiv.org/abs/1408.5093 (accessed November 22, 2022)
- Google Colab homepage. Available at: https://colab. research.google.com (accessed November 22, 2022)
- Lavin, Scott Gray. 2015. Fast Algorithms for Convolutional Neural Networks. arXiv:1509.09308. Available at: https://arxiv.org/ abs/1509.09308 (accessed November 22, 2022)
- Anton V. Trusov, Elena E. Limonova, Dmitry P. Nikolaev and Vladimir V. Arlazarov. 2021. p-im2col: Simple Yet Efficient Convolution Algorithm With Flexibly Controlled Memory Overhead. IEEE Access PP(99):1-1 (2021)
Arquivos suplementares
