Online Adaptation for Implicit Object Tracking and Shape Reconstruction in the Wild
This work addresses a key problem for computer vision, robotics, and autonomous driving systems by enhancing 3D perception in real-world, cluttered environments, though it is incremental as it builds on existing implicit function models like DeepSDF.
The paper tackles the challenge of generalizing 3D object tracking and shape reconstruction to cluttered, partially observable LiDAR data by leveraging video continuity, proposing a unified framework that adapts a neural implicit function online to iteratively improve both tasks, resulting in significant improvements over state-of-the-art methods on Waymo and KITTI datasets.
Tracking and reconstructing 3D objects from cluttered scenes are the key components for computer vision, robotics and autonomous driving systems. While recent progress in implicit function has shown encouraging results on high-quality 3D shape reconstruction, it is still very challenging to generalize to cluttered and partially observable LiDAR data. In this paper, we propose to leverage the continuity in video data. We introduce a novel and unified framework which utilizes a neural implicit function to simultaneously track and reconstruct 3D objects in the wild. Our approach adapts the DeepSDF model (i.e., an instantiation of the implicit function) in the video online, iteratively improving the shape reconstruction while in return improving the tracking, and vice versa. We experiment with both Waymo and KITTI datasets and show significant improvements over state-of-the-art methods for both tracking and shape reconstruction tasks. Our project page is at https://jianglongye.com/implicit-tracking .