Abstract
We report on experiences with optimizations that enable neural network edge computing on a small robot vehicle, which serves as a proxy for an autonomous robotic space craft. We realize a visual object recognition task using a neural network on a field-programmable gate array (FPGA) with data processing resources as limited as those of an FPGA suitable for space. We use a quantized neural network nicely matching the properties of an FPGA. The restrictions of the small FPGA require us to sequentialize the processing partially. We employ input frame tiling for this. It allows us to keep the entire neural network on-chip. Furthermore, we split up the visual object recognition task into two stages, using two separate neural networks. The first stage identifies the region of interest approximately, using large and thus few tiles. The second stage looks closely at the single tile containing the region of interest; thus being not that time critical.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Altnickel, P., Catalkaya, T., Bredereke, J., et al.: Gesten- und Objekterkennung durch schwache FPGAs in autonomen Fahrzeugen mittels neuronaler Netze. German. Technical report. HSB (2021)
Altnickel, P., Cansu, F., Bredereke, J., et al.: Personenerkennung durch schwache FPGAs in autonomen Fahrzeugen mittels Neuronaler Netze. German. Technical report. HSB (2022)
Blott, M., et al.: FINN-R: an end-to-end deep-learning framework for fast exploration of quantized neural networks. ACM Trans. Reconfigurable Technol. Syst. 11(3) (2018). https://doi.org/10.1145/3242897
von Huth, C., Soldin, M., Bredereke, J., et al.: Bericht zum Projekt “Neuronale Netze auf strahlungstoleranten FPGAs für die Raumfahrt”. German. Technical report. HSB (2020)
Jokic, P., et al.: BinaryEye: a 20 kfps streaming camera system on FPGA with real-time on-device image recognition using binary neural networks. In: 2018 IEEE 13th International Symposium on Industrial Embedded Systems (SIES) (2018). https://doi.org/10.1109/SIES.2018.8442108
Manning, J., et al.: Machine-learning space applications on SmallSat platforms with TensorFlow. In: Proceedings of AIAA/USU Conference on Small Satellites (SmallSat) (2018)
Müller, F., Krekel, N., Bredereke, J., et al.: Applying binarized neural networks on FPGAs to an autonomous driving problem. Technical report. HSB (2021)
Müller, F.: Dynamisches Tiling auf schwachen FPGAs zur Objekterkennung mithilfe kleiner neuronaler Netze. German. Bachelorthesis. HSB (2021)
Roffe, S., et al.: CASPR: autonomous sensor processing experiment for STP-H7. In: Small Satellite Conference (2020)
Umuroglu, et al.: FINN: a framework for fast, scalable binarized neural network inference. In: Proceedings of the 2017 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays, FPGA 2017, pp. 65–74. ACM (2017). https://doi.org/10.1145/3020078.3021744
Xu, X., et al.: DAC-SDC low power object detection challenge for UAV applications. IEEE Trans. Pattern Anal. Mach. Intell. 43(2), 392–403 (2021). https://doi.org/10.1109/TPAMI.2019.2932429
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Bredereke, J. (2023). Enabling Neural Network Edge Computing on a Small Robot Vehicle. In: Braubach, L., Jander, K., Bădică, C. (eds) Intelligent Distributed Computing XV. IDC 2022. Studies in Computational Intelligence, vol 1089. Springer, Cham. https://doi.org/10.1007/978-3-031-29104-3_4
Download citation
DOI: https://doi.org/10.1007/978-3-031-29104-3_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-29103-6
Online ISBN: 978-3-031-29104-3
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)