SeqTrack3D: Exploring Sequence Information for Robust 3D Point Cloud Tracking

3D single object tracking (SOT) is an important and challenging task for the autonomous driving and mobile robotics. Most existing methods perform tracking between two consecutive frames while ignoring the motion patterns of the target over a series of frames, which would cause performance degradation in the scenes with sparse points. To break through this limitation, we introduce "Sequence-to-Sequence" tracking paradigm and a tracker named SeqTrack3D to capture target motion across continuous frames. Unlike previous methods that primarily adopted three strategies: matching two consecutive point clouds, predicting relative motion, or utilizing sequential point clouds to address feature degradation, our SeqTrack3D combines both historical point clouds and bounding box sequences. This novel method ensures robust tracking by leveraging location priors from historical boxes, even in scenes with sparse points. Extensive experiments conducted on large-scale datasets show that SeqTrack3D achieves new state-of-the-art performances, improving by 6. 00% on NuScenes and 14. 13% on Waymo dataset. The code will be made public at https://github.com/aron-lin/seqtrack3d.

Authors

• Yu Lin
• Zhiheng Li 0003
• Yubo Cui
• Zheng Fang 0001

Keywords

• Point cloud compression
• Degradation
• Target tracking
• Three-dimensional displays
• Codes
• Transformers
• Object tracking
• Point Cloud
• Robust Tracking
• Bounding Box
• Motion Patterns
• Single Tracking
• Sparse Point
• Box Sequence
• Series Of Frames
• Transformer
• Computer Vision
• Spatial Features
• Object Detection
• Recurrent Neural Network
• Attention Mechanism
• Temporal Features
• Intersection Over Union
• Multilayer Perceptron
• Feed-forward Network
• Feature Points
• Area Under Curve
• Length Of The Time Window
• Continuous Motion
• Current Frame
• Sequence Learning
• Local Coordinate System
• Historical Framing
• Point In Frame