Accurate and Real-time Pseudo Lidar Detection: Is Stereo Neural Network Really Necessary?
This work addresses real-time deployment for autonomous vehicles by making Pseudo-Lidar systems more efficient, though it is incremental as it builds on existing representation methods.
The paper tackles the problem of high latency in Pseudo-Lidar-based 3D object detection by showing that fast but inaccurate stereo depth estimation can achieve competitive accuracy with proper refinement, achieving 23 ms computing time on the KITTI benchmark.
The proposal of Pseudo-Lidar representation has significantly narrowed the gap between visual-based and active Lidar-based 3D object detection. However, current researches exclusively focus on pushing the accuracy improvement of Pseudo-Lidar by taking the advantage of complex and time-consuming neural networks. Seldom explore the profound characteristics of Pseudo-Lidar representation to obtain the promoting opportunities. In this paper, we dive deep into the pseudo Lidar representation and argue that the performance of 3D object detection is not fully dependent on the high precision stereo depth estimation. We demonstrate that even for the unreliable depth estimation, with proper data processing and refining, it can achieve comparable 3D object detection accuracy. With this finding, we further show the possibility that utilizing fast but inaccurate stereo matching algorithms in the Pseudo-Lidar system to achieve low latency responsiveness. In the experiments, we develop a system with a less powerful stereo matching predictor and adopt the proposed refinement schemes to improve the accuracy. The evaluation on the KITTI benchmark shows that the presented system achieves competitive accuracy to the state-of-the-art approaches with only 23 ms computing, showing it is a suitable candidate for deploying to real car-hold applications.