@FinanceYF5: A Chinese open-source model can reconstruct 3D scenes in real-time from any video. Just one camera, no LiDAR needed, handles over 10,000 frames without crashing, and runs at 20fps on a single GPU. Its benchmark scores surpass traditional optimization-based methods, tested on drone aerial photography, dashcams, and indoor tours. And it's fully open-source.
Summary
A Chinese open-source model reconstructs 3D scenes in real-time from a single camera video, no LiDAR required, achieves 20fps on a single GPU with performance superior to traditional optimization methods. Fully open-source.
View Cached Full Text
Cached at: 07/20/26, 09:28 AM
A domestic open-source model can reconstruct 3D scenes in real-time from any short video clip.
Just a single camera, no LiDAR needed. It handles over 10,000 frames without crashing, and runs at 20fps on a single GPU. Its benchmark scores outperform traditional optimization-based methods, and it has been tested on drone aerial footage, dashcam recordings, and indoor walkthroughs.
And it’s fully open-source. https://t.co/6LHp8Npkkz
Similar Articles
@thesupermanmx: China open-sourced a model that reconstructs any scene in 3D from a regular video, in real-time. one camera. no LiDAR. …
China open-sourced a real-time 3D scene reconstruction model that works from a single regular video without LiDAR, achieving 20 FPS on a single GPU and maintaining stability over 10,000+ frames.
@FinanceYF5: This AI is impressive. LingBot-Map can convert real-time video streams into real-time 3D reconstruction. 20 FPS code + model
LingBot-Map is an AI model capable of converting real-time video streams into real-time 3D reconstruction, running at 20 FPS with complete code and model provided.
@XAMTO_AI: ControlNet author Min Shen has come up with something new! The newly open-sourced FramePack directly lowers the barrier for video generation — runs on just 6GB VRAM, generates a 1-minute 30fps video with a 13B model, and on an RTX 4090 it takes only 1.5 seconds per frame. Such configuration requirements were unimaginable before. The core idea is frame-by-frame…
ControlNet author Min Shen has open-sourced the FramePack video generation model, which requires only 6GB of VRAM to run a 13B model, generates a 1-minute 30fps video, takes 1.5 seconds per frame on an RTX 4090, and comes with a one-click Windows package.
@IlirAliu_: Forget lidar. One single camera. Runs in real time & is open source: A streaming 3D model that reconstructs scenes live…
LingBot-Map is an open-source, real-time streaming 3D reconstruction model that uses a single camera, running at ~20 FPS via a feed-forward geometric context transformer, outperforming both streaming and offline methods.
@FinanceYF5: 1/ Most AI "world models" suffer from visual collapse and scene drift after just a few minutes of running. @robbyant_brain just open-sourced LingBot-World 2.0 (Infinity): supports hour-level real-time generation, no quality degradation over long runs, stable 720p/60fps output. This is...
LingBot-World 2.0 (Infinity) has just been open-sourced, supporting hour-level real-time generation with no quality degradation over long runs and stable 720p/60fps output. It's a key step for robots to learn to predict the world.