UPAQ: A Framework for Real-Time and Energy-Efficient 3D Object Detection in Autonomous Vehicles
This work addresses efficiency challenges for autonomous vehicles by enabling real-time and energy-efficient 3D object detection on resource-constrained embedded platforms, representing an incremental improvement over existing compression frameworks.
The paper tackles the problem of high memory and computational costs in 3D object detection for autonomous vehicles by introducing UPAQ, a framework using semi-structured pattern pruning and quantization, which achieves up to 5.62x model compression, 1.97x inference speed boost, and 2.07x energy reduction compared to state-of-the-art methods.
To enhance perception in autonomous vehicles (AVs), recent efforts are concentrating on 3D object detectors, which deliver more comprehensive predictions than traditional 2D object detectors, at the cost of increased memory footprint and computational resource usage. We present a novel framework called UPAQ, which leverages semi-structured pattern pruning and quantization to improve the efficiency of LiDAR point-cloud and camera-based 3D object detectors on resource-constrained embedded AV platforms. Experimental results on the Jetson Orin Nano embedded platform indicate that UPAQ achieves up to 5.62x and 5.13x model compression rates, up to 1.97x and 1.86x boost in inference speed, and up to 2.07x and 1.87x reduction in energy consumption compared to state-of-the-art model compression frameworks, on the Pointpillar and SMOKE models respectively.