Abstract
Deep learning is moving more and more from the cloud towards the edge. Therefore, embedded devices are needed that are reasonably cheap, energy-efficient and fast enough. In this paper we evaluate the performance and energy consumption of popular, off-the-shelf commercial devices for deep learning inferencing. We compare the Intel Neural Compute Stick 2, the Google Coral Edge TPU and the Nvidia Jetson Nano with the Raspberry Pi 4 for their suitability as a central controller in an autonomous vehicle for the formula student driverless.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
- 2.
- 3.
- 4.
Previous measurements have shown that the type of background and the number of detected objects has no measurable influence on the operating time and energy consumption. We’ve published our dataset on https://github.com/iisys-hof/formula-student-dataset.
References
Liu, L., Ouyang, W., Wang, X., Fieguth, P., Chen, J., Liu, X., Pietikäinen, M.: Deep learning for generic object detection: a survey. Int. J. Comput. Vis. 128, 261–318 (2020)
Storks, S., Gao, Q., Chai, J.Y.: Recent Advances in Natural Language Inference: A Survey of Benchmarks, Resources, and Approaches. ArXiv Prepr. ArXiv190401172 (2019)
Chen, J., Ran, X.: Deep learning with edge computing: A review. Proc. IEEE 107, 1655–1674 (2019)
Srinivasa, S.S., et al.: MuSHR: A Low-Cost, Open-Source Robotic Racecar for Education and Research. ArXiv Prepr. ArXiv190808031 (2019)
Zeilinger, M., Hauk, R., Bader, M., Hofmann, A.: Design of an autonomous race car for the formula student driverless (fsd). In: Oagm & Arw Joint Workshop (2017)
Jun, N.I., Jibin, H.U.: Autonomous driving system design for formula student driverless racecar. In: 2018 IEEE Intelligent Vehicles Symposium (IV). pp. 1–6. IEEE (2018)
Kabzan, J., et al.: Amz driverless: The full autonomous racing system. ArXiv Prepr. ArXiv190505150 (2019)
Bahl, G., Daniel, L., Moretti, M., Lafarge, F.: Low-power neural networks for semantic segmentation of satellite images. In: Proceedings of the IEEE International Conference on Computer Vision Workshops (2019)
Kristiani, E., Yang, C.-T., Huang, C.-Y.: iSEC: an optimized deep learning model for image classification on edge computing. IEEE Access. 8, 27267–27276 (2020)
He, Q., Weaver, V., Segee, B.: Comparing power and energy usage for scientific calculation with and without GPU acceleration on a raspberry Pi model B + and 3B. In: International Conference on Internet Computing (ICOMP), pp. 3–9 (2018)
Peng, T., Zhang, D., Hettiarachchi, D.L.N., Loomis, J.: An evaluation of embedded GPU systems for visual SLAM algorithms. In: IS&T International Symposium on Electronic Imaging 2020: Intelligent Robotics and Industrial Applications using Computer Vision (2020)
Pinto, D.: NVIDIA Releases Jetson Xavier NX Developer Kit with Cloud-Native Support. http://nvidianews.nvidia.com/news/nvidia-releases-jetson-xavier-nx-developer-kit-with-cloud-native-support. Accessed 23 May 2020
Lee, M., Hwang, K., Park, J., Choi, S., Shin, S., Sung, W.: FPGA-based low-power speech recognition with recurrent neural networks. In: 2016 IEEE International Workshop on Signal Processing Systems (SiPS), pp. 230–235. IEEE (2016)
Reagen, B., et al.: Minerva: enabling low-power, highly-accurate deep neural network accelerators. In: 2016 ACM/IEEE 43rd Annual Intl. Symposium on Computer Architecture (ISCA), pp. 267–278. IEEE (2016)
Libutti, L.A., Igual, F.D., Pinuel, L., De Giusti, L., Naiouf, M.: Benchmarking performance and power of USB accelerators for inference with MLPerf⋆. In: 2nd Workshop on Accelerated Machine Learning (AccML), Valencia, Spain (2020)
Hossain, S., Lee, D.: Deep learning-based real-time multiple-object detection and tracking from aerial imagery via a flying robot with gpu-based embedded devices. Sensors. 19, 3371 (2019)
Chen, C.-W., Ruan, S.-J., Lin, C.-H., Hung, C.-C.: Performance evaluation of edge computing-based deep learning object detection. In: 7th International Conference on Network, Communication and Computing, pp. 40–43 (2018)
Hegde, G., Kapre, N.: CaffePresso: accelerating convolutional networks on embedded SoCs. ACM Trans. Embed. Comput. Syst. TECS. 17, 1–26 (2017)
Gao, C., Rios-Navarro, A., Chen, X., Delbruck, T., Liu, S.-C.: EdgeDRNN: Enabling Low-latency Recurrent Neural Network Edge Inference. ArXiv Prepr. ArXiv191212193 (2019)
Gorbachev, Y., Fedorov, M., Slavutin, I., Tugarev, A., Fatekhov, M., Tarkan, Y.: OpenVINO Deep Learning Workbench: Comprehensive Analysis and Tuning of Neural Networks Inference. In: IEEE Intl. Conf. on Computer Vision Workshops. (2019)
Du, Z., et al.: ShiDianNao: Shifting vision processing closer to the sensor. In: 42nd Annual International Symposium on Computer Architecture, pp. 92–104 (2015)
Chen, Y., Chen, T., Xu, Z., Sun, N., Temam, O.: DianNao family: energy-efficient hardware accelerators for machine learning. Commun. ACM 59, 105–112 (2016)
Diehl, P.U., Zarrella, G., Cassidy, A., Pedroni, B.U., Neftci, E.: Conversion of artificial recurrent neural networks to spiking neural networks for low-power neuromorphic hardware. In: 2016 IEEE International Conference on Rebooting Computing (ICRC), pp. 1–8. IEEE (2016)
Lane, N.D., et al..: Deepx: a software accelerator for low-power deep learning inference on mobile devices. In: 15th ACM/IEEE International Conference on Info Proceeding in Sensor Nets (IPSN). pp. 1–12. IEEE (2016)
Peinl, R., Rizk, B., Szabad, R.: Open source speech recognition on edge devices. In: 10th International Conference on Advanced Computer Information Technologies, Deggendorf, Germany (2020)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Puchtler, P., Peinl, R. (2020). Evaluation of Deep Learning Accelerators for Object Detection at the Edge. In: Schmid, U., Klügl, F., Wolter, D. (eds) KI 2020: Advances in Artificial Intelligence. KI 2020. Lecture Notes in Computer Science(), vol 12325. Springer, Cham. https://doi.org/10.1007/978-3-030-58285-2_29
Download citation
DOI: https://doi.org/10.1007/978-3-030-58285-2_29
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-58284-5
Online ISBN: 978-3-030-58285-2
eBook Packages: Computer ScienceComputer Science (R0)