GitHub Trending · AI Apps & Models · Updated
Robbyant/lingbot-map
A geometric context transformer foundation model for streaming 3D reconstruction
- Stars
- ★ 17.7k
- Language
- Python
- Chart appearances
- 3
- On-chart tracking
- 1 days tracked
What it does
LingBot-Map is a feed-forward 3D foundation model focused on streaming 3D reconstruction. Its Geometric Context Transformer unifies coordinate grounding, dense geometric cues, and long-range drift correction within a single streaming framework for efficient, stable online reconstruction.
What's inside
The repo contains inference code, an interactive demo.py, an offline rendering pipeline in demo_render/batch_demo.py, plus installation instructions, model download links, and evaluation benchmark scripts.
Tech stack
Implemented in Python with PyTorch and attention backends (FlashInfer / SDPA), using paged KV cache attention to support long-sequence inference.
Use cases
- 3D reconstruction and SLAM researchers for streaming reconstruction and long-sequence drift correction experiments
- Robotics/AR-VR developers needing real-time online 3D perception and mapping
- Computer vision engineers evaluating feed-forward 3D foundation models on benchmarks
Why it's trending
As an ECCV 2026 Best Paper Award candidate, it quickly gained 17,703 stars thanks to ~20 FPS streaming inference and stable reconstruction over 10,000+ frames, adding 110 stars today and ranking 10th on GitHub Trending daily.
Current chart placements
| Period | Language filter | Rank | Stars gained |
|---|---|---|---|
| Daily | All languages | #10 | +110 |
| Daily | Python | #3 | +110 |
| Weekly | Python | #7 | +446 |
Editorial summaries are based on the public repository and the dated GitHub Trending snapshot. Verify implementation details in the upstream repository.