Journal of the Korea Institute of Information and Communication Engineering (한국정보통신학회논문지)

The Korea Institute of Information and Commucation Engineering (한국정보통신학회)
권호 표지
DOI QR코드

DOI QR Code

Priority-based Multi-DNN scheduling framework for autonomous vehicles

자율주행차용 우선순위 기반 다중 DNN 모델 스케줄링 프레임워크

  • Cho, Ho-Jin (Department of IT Convergence Engineering, Hansung University) ;
  • Hong, Sun-Pyo (Department of IT Convergence Engineering, Hansung University) ;
  • Kim, Myung-Sun (Department of IT Convergence Engineering, Hansung University)
  • Received : 2021.01.12
  • Accepted : 2021.02.05
  • Published : 2021.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

With the recent development of deep learning technology, autonomous things technology is attracting attention, and DNNs are widely used in embedded systems such as drones and autonomous vehicles. Embedded systems that can perform large-scale operations and process multiple DNNs for high recognition accuracy without relying on the cloud are being released. DNNs with various levels of priority exist within these systems. DNNs related to the safety-critical applications of autonomous vehicles have the highest priority, and they must be handled first. In this paper, we propose a priority-based scheduling framework for DNNs when multiple DNNs are executed simultaneously. Even if a low-priority DNN is being executed first, a high-priority DNN can preempt it, guaranteeing the fast response characteristics of safety-critical applications of autonomous vehicles. As a result of checking through extensive experiments, the performance improved by up to 76.6% in the actual commercial board.

최근 딥러닝 기술이 발전함에 따라 자율 사물 기술이 주목받으면서 드론이나 자율주행차 같은 임베디드 시스템에서 DNN을 많이 활용하고 있다. 클라우드에 의지하지 않고 높은 인식 정확도를 위해서 큰 규모의 연산이 가능하고 다수의 DNN을 처리할 수 있는 임베디드 시스템들이 출시되고 있다. 이러한 시스템 내부에는 다양한 수준의 우선순위를 갖는 DNN들이 존재한다. 자율주행차의 안전 필수에 관련된 DNN들은 가장 높은 우선순위를 갖고 이들은 반드시 최우선적으로 처리되어야 한다. 본 논문에서는 다수의 DNN이 동시에 실행될 때 우선순위를 고려해서 DNN을 스케줄링하는 프레임워크를 제안한다. 낮은 우선순위의 DNN이 먼저 실행되고 있어도 높은 우선순위의 DNN이 이를 선점할 수 있어 자율주행차의 안전 필수 응용의 빠른 응답 특성을 보장한다. 실험을 통하여 확인한 결과 실제 상용보드에서 최대 76.6% 성능이 향상되었다.

Keywords

References

  1. J. Gao, J. Du, and E. Chen, "Mixed-Bandwidth CrossChannel Speech Recognition via Joint Optimization of DNN-Based Bandwidth Expansion and Acoustic Modeling," in IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 27, no. 3, pp. 559-571, Mar. 2019. DOI: 10.1109/TASLP.2018.2886739.
  2. K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," in Proceedings of the International Conference on Learning Representations, San Diego, CA, 2015.
  3. S. Liu and Z. Liu, "Multi-channel CNN-based object detection for enhanced situation awareness," arXiv preprint arXiv:1712.00075, 2017.
  4. A. Krizhevsky, I. Sutskever, and G. E. Hinton, "ImageNet classification with deep convolutional neural networks," in Proceedings of the 26th Conference on Neural Information Processing Systems, Lake Tahoe, pp. 1097-1105, 2012.
  5. H. Kim, J. Kim, and H. Jung, "Convolutional Neural Network Based Image Processing System," Journal of Information and Communication Convergence Engineering, vol. 16, no. 3, pp. 160-165, Sep. 2018. https://doi.org/10.6109/JICCE.2018.16.3.160
  6. K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, pp. 770-778, 2016.
  7. H. Li, D. Yu, A. Kumar, and Y. -C. Tu, "Performance modeling in CUDA streams -A means for high-throughput data processing," 2014 IEEE International Conference on Big Data (Big Data), Washington, DC, pp. 301-310, 2014. DOI: 10.1109/BigData.2014.7004245.
  8. HyperQ [Internet]. Available: http://developer.download.nvidia.com/compute/DevZone/C/html_x64/6_Advanced/simpleHyperQ/doc/HyperQ.pdf.
  9. CUDA API [Internet]. Available: https://docs.nvidia.com/cuda/.
  10. T. Amert, N. Otterness, M. Yang, J. H. Anderson, and F. D. Smith, "GPU Scheduling on the NVIDIA TX2: Hidden Details Revealed," 2017 IEEE Real-Time Systems Symposium (RTSS), Paris, pp. 104-115, 2017. DOI: 10.1109/RTSS.2017.00017.
  11. Y. Xiang and H. Kim, "Pipelined Data-Parallel CPU/GPU Scheduling for Multi-DNN Real-Time Inference," 2019 IEEE Real-Time Systems Symposium (RTSS), Hong Kong, China, pp. 392-405, 2019. DOI: 10.1109/RTSS46320.2019.00042.
  12. D. Kim and M. Kim "Avoiding Performance Degradation of DNN based Safety-Critical Applications in Autonomous Vehicles," Journal of the Institute of Electronics and Information Engineers, vol. 57, no. 9, pp. 56-64, Sep. 2019. https://doi.org/10.5573/ieie.2020.57.9.56
  13. Jetson AGX Xavier Developer Kit [Internet]. Available: https://developer.nvidia.com/embedded/jetson-agx-xavier-developer-kit.
  14. X. Yu, N. Zeng, S. Liu, and Y. D. Zhang, "Utilization of DenseNet201 for diagnosis of breast abnormality," in Machine Vision and Applications, vol. 30, Oct. 2019.
  15. L. Nguyen, D. Lin, Z. Lin, and J. Cao, "Deep CNNs for microscopic image classification by exploiting transfer learning and feature concatenation," in Proceedings of the IEEE International Symposium on Circuits and Systems (ISCAS), Florence, Italy, 2018.
  16. NVIDIA Visual Profiler [Internet]. Available: https://developer.nvidia.com/nvidia-visual-profiler.