wenbowen123/BundleTrack
A C++ research implementation of a deep learning-based 6D pose tracking system for novel objects in robotics manipulation scenarios.
Not currently ranked — collecting fresh signals.
star history
BundleTrack performs 6D pose tracking of novel objects without requiring CAD models by combining deep learning-based segmentation and feature extraction with memory-augmented pose graph optimization for spatiotemporal consistency. The system handles challenging scenarios including significant occlusions and object motions, achieving real-time performance at 10Hz via CUDA implementation.
Frequently asked
- What is wenbowen123/BundleTrack?
- A C++ research implementation of a deep learning-based 6D pose tracking system for novel objects in robotics manipulation scenarios.
- Is BundleTrack open source?
- Yes — wenbowen123/BundleTrack is open source, released under the Apache-2.0 license.
- What language is BundleTrack written in?
- wenbowen123/BundleTrack is primarily written in C++.
- How popular is BundleTrack?
- wenbowen123/BundleTrack has 686 stars on GitHub.
- Where can I find BundleTrack?
- wenbowen123/BundleTrack is on GitHub at https://github.com/wenbowen123/BundleTrack.