Journal
MACHINE LEARNING-SCIENCE AND TECHNOLOGY
Volume 3, Issue 4, Pages -Publisher
IOP Publishing Ltd
DOI: 10.1088/2632-2153/ac9cb5
Keywords
FPGA; computer vision; deep learning; hls4ml; machine learning; autonomous vehicles; semantic segmentation
Categories
Funding
- European Research Council (ERC) [772369, 966696]
- European Research Council (ERC) [772369, 966696] Funding Source: European Research Council (ERC)
Ask authors/readers for more resources
This paper investigates the use of field programmable gate arrays as hardware accelerators for real-time semantic segmentation tasks in autonomous driving. By exploring the deployment and implementation optimization, a fully-on-chip deployment of the compressed ENet convolutional neural network architecture is achieved on a Xilinx ZCU102 evaluation board, resulting in reduced latency and resource utilization.
In this paper, we investigate how field programmable gate arrays can serve as hardware accelerators for real-time semantic segmentation tasks relevant for autonomous driving. Considering compressed versions of the ENet convolutional neural network architecture, we demonstrate a fully-on-chip deployment with a latency of 4.9 ms per image, using less than 30% of the available resources on a Xilinx ZCU102 evaluation board. The latency is reduced to 3 ms per image when increasing the batch size to ten, corresponding to the use case where the autonomous vehicle receives inputs from multiple cameras simultaneously. We show, through aggressive filter reduction and heterogeneous quantization-aware training, and an optimized implementation of convolutional layers, that the power consumption and resource utilization can be significantly reduced while maintaining accuracy on the Cityscapes dataset.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available