Neural Pruning Search for Real-Time Object Detection of Autonomous Vehicles

Object detection plays an important role in self-driving cars for security development. However, mobile systems on self-driving cars with limited computation resources lead to difficulties for object detection. To facilitate this, we propose a compiler-aware neural pruning search framework to achiev...

Full description

Saved in:
Bibliographic Details
Published in:2021 58th ACM/IEEE Design Automation Conference (DAC) pp. 835 - 840
Main Authors: Zhao, Pu, Yuan, Geng, Cai, Yuxuan, Niu, Wei, Liu, Qi, Wen, Wujie, Ren, Bin, Wang, Yanzhi, Lin, Xue
Format: Conference Proceeding
Language:English
Published: IEEE 05.12.2021
Subjects:
Autonomous automobiles
autonomous driving
Design automation
Mobile handsets
Object detection
real-time
Real-time systems
Search problems
Three-dimensional displays
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Object detection plays an important role in self-driving cars for security development. However, mobile systems on self-driving cars with limited computation resources lead to difficulties for object detection. To facilitate this, we propose a compiler-aware neural pruning search framework to achieve high-speed inference on autonomous vehicles for 2D and 3D object detection. The framework automatically searches the pruning scheme and rate for each layer to find a best-suited pruning for optimizing detection accuracy and speed performance under compiler optimization. Our experiments demonstrate that for the first time, the proposed method achieves (close-to) real-time, 55ms and 97ms inference times for YOLOv4 based 2D object detection and PointPillars based 3D detection, respectively, on an off-the-shelf mobile phone with minor (or no) accuracy loss.
DOI:10.1109/DAC18074.2021.9586163