Electronics and Telecommunications Trends (전자통신동향분석)

Electronics and Telecommunications Research Institute (한국전자통신연구원)
권호 표지
DOI QR코드

DOI QR Code

Trend of Edge Machine Learning as-a-Service

서비스형 엣지 머신러닝 기술 동향

  • Published : 2022.10.01
Download PDF
⟨ Previous Next ⟩

Abstract

The Internet of Things (IoT) is growing exponentially, with the number of IoT devices multiplying annually. Accordingly, the paradigm is changing from cloud computing to edge computing and even tiny edge computing because of the low latency and cost reduction. Machine learning is also shifting its role from the cloud to edge or tiny edge according to the paradigm shift. However, the fragmented and resource-constrained features of IoT devices have limited the development of artificial intelligence applications. Edge MLaaS (Machine Learning as-a-Service) has been studied to easily and quickly adopt machine learning to products and overcome the device limitations. This paper briefly summarizes what Edge MLaaS is and what element of research it requires.

Keywords

Acknowledgement

이 논문은 2022년도 정부(과학기술정보통신부)의 재원으로 정보통신기획평가원의 지원을 받아 수행된 연구임[No. 2022-0-00454, 스마트 엣지 디바이스 SW 개발 플랫폼 개발].

References

  1. P.P. Ray, "A review on TinyML: State-of-the-art and prospects," J. King Saud Univ.: Comput. Inf. Sci., vol. 34, no. 4, 2021.
  2. S. Soro, "TinyML for ubiquitous edge AI," arXiv preprint, CoRR, 2021, arXiv: 2102.01255.
  3. V. Rajapakse, I. Karunanayake, and N. Ahmed, "Intelligence at the extreme edge: A survey on reformable TinyML," arXiv preprint, CoRR, 2022, arXiv: 2204.00827.
  4. R. Sanchez Iborra and A.F. Skarmeta, "TinyML-Enabled frugal smart objects challenges and opportunities," IEEE Circuits Syst. Mag., vol. 20, no. 3, 2020, pp. 4-8. https://doi.org/10.1109/mcas.2020.3005467
  5. https://www.tinyml.org/
  6. 신성필, "서비스형 기계학습(MLaaS, Machine Learning as a Service) 시장 동향 및 기능 요구사항 표준," TTA저널, 제198호 제11/12월호, 2021.
  7. H. Doyu et al., "A tinymlaas ecosystem for machine learning in iot: Overview and research challenges," in Proc. Int. Symp. VLSI Des., Autom. Test (VLSI-DAT), (Hsinchu, Taiwan), Apr. 2021.
  8. C. Rogers, "The opportunity for AI at the edge and beyond," SensiML, Apr. 2020.
  9. 김대우, "2021년 머신러닝과 인공지능(AI) 트렌드-MLaaS," Jan. 2021, https://sqlermail.medium.com
  10. Vertiv, Data Center 2025-Closer To The Edge, 2019, https://www.vertiv.com/en-emea/about/news-and-insights/articles/pr-campaigns-reports/data-center-2025-closer-to-the-edge/
  11. https://aws.amazon.com/ko/sagemaker/edge/
  12. https://cloud.google.com/vertex-ai
  13. https://docs.microsoft.com/ko-kr/azure/iot-edge/
  14. https://octoml.ai/
  15. https://www.edgeimpulse.com/
  16. https://sensiml.com/
  17. Google Inc., TensorFlow Lite, https://www.tensorflow.org/lite/
  18. Google Inc., TensorFlow Lite for Microcontrollers, https://www.tensorflow.org/lite/microcontrollers
  19. Microsoft Corp., Embedded Learning Library, https://microsoft.github.io/ELL/
  20. ARM In., ARM-NN, https://www.arm.com/products/silicon-ip-cpu/ethos/arm-nn
  21. https://autokeras.com
  22. https://www.tiriasresearch.com/wp-content/uploads/2021/09/Edge-Impulse-Accelerates-MLOps-with-EON.pdf
  23. https://github.com/microsoft/nni
  24. S. Leroux et al., "TinyMLOps: Operational challenges for widespread edge AI Adoption," arXiv preprint, CoRR, 2022, arXiv: 2203.10923v2.
  25. M. Rastegari et al., "XnorNet: ImageNet classification using binary convolutional neural networks," arXiv preprint, CoRR, 2016, arXiv: 1603.05279.
  26. W. Chen, P. Wang, and J. Cheng, "Towards mixed-precision quantization o f neural networks via constrained cptimization," arXiv preprint, CoRR, 2021, arXiv: 2110.06554.
  27. K. Wang et al., "HanHAQ: Hardware-aware automated quantization," arXiv preprint, CoRR, 2019, arXiv: 1811.08886.
  28. Y. He et al., "Filter pruning via geometric median for deep convolutional neural networks acceleration," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit., (Long Beach, CA, USA), June 2019, pp. 4340-4349.
  29. Y. He et al., "Amc: Automl for model compression and acceleration on mobile devices," in Proc. Eur. Conf. Comput. Vis. (ECCV), (Munich, Germany), Sept. 2018, pp. 784-800.
  30. J. Martinez et al., "Permute, quantize, and fine-tune: Efficient compression of neural networks," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit., (Virtual), June 2021, pp. 15699-15708.
  31. T. Kim et al., "CPrune: Compiler-informed model pruning for efficient target-aware DNN execution," arXiv preprint, CoRR, 2022, arXiv: 2207.01260.
  32. T.J. Yang et al., "Netadapt: Platform-Aware neural network adaptation for mobile applications," in Proc. Eur. Conf. Comput. Vis. (ECCV), (Munich, Germany), Sept. 2018, pp. 285-300.
  33. K.T. Chitty-Venkata and A.K. Somani, "Neural architecture search survey: A hardware perspective," ACM Comput. Surv., 2022, doi: 10.1145/3524500.
  34. https://github.com/etri/nest-compiler
  35. X. Team, Xla: Domain-Specific Compiler for Linear Algebra that Optimizes Tensorflow Computations, 2019.
  36. W.F. Lin et al., "Onnc: A compilation framework connecting onnx to proprietary deep learning accelerators," in Proc. IEEE Int. Conf. Artif. Intell. Circuits Sys. (AICAS), (Hsinchu, Taiwan), Mar. 2019, pp. 214-218.
  37. S. Cyphers et al., "Intel nGraph: An intermediate representation, compiler, and executor for deep learning," arXiv preprint, CoRR, 2018, arXiv: 1801.08058.
  38. N. Rotem et al., "Glow: Graph lowering compiler techniques for neural networks," arXiv preprint, CoRR, 2018, arXiv: 1805.00907.
  39. T. Chen et al., "Tvm: An automated end-to-end optimizing compiler for deep learning," in Proc. USENIX Symp. Oper. Syst. Des. Implement., (Carlsbad, CA, USA), Oct. 2018, pp. 578-594.
  40. M. Li et al., "The deep learning compiler: A comprehensive survey," arXiv preprint, CoRR, 2020, arXiv: 2002.03794v4. https://doi.org/10.1109/TPDS.2020.3030548
  41. R. David et al., "TensorFlow lite micro: Embedded machine learning for TinyML systems," in Proc. Mach. Lear. Syst., (San Jose, CA, USA), 2021.
  42. Microsoft Corp., The Embedded Learning Library-Embedded Learning Library(ELL), 2022, Retrieved from https://microsoft.github.io/ELL/
  43. W. Li and M. Liewig, "A survey of AI accelerators for edge environment," in World Conference on Information Systems and Technologies, Springer, Cham, Switzerland, 2020, pp. 35-44.