正在加载视频...
视频加载失败
Monocular visual slam with loop closures on a $10 chip. cc Akshat
17,622 次观看 • 23 天前 •via X (Twitter)
16 条评论
@akshat_sj hell yeah 🦾
@akshat_sj That’s really impressive, sitting with slam rn for my autonomous drone
@akshat_sj Neat. Which chip did you go with.
@akshat_sj 3566!
@akshat_sj Isaac vslam barely runs good in jetson 8gb idk what im doing wrong
@akshat_sj Most slam systems are not optimised at all
@akshat_sj Fun! This looks so good, what resolution and fps?
@akshat_sj It's QVGA. Gives stable 15 fps.
@akshat_sj Which chip?
@akshat_sj Is it pure VSLAM or VISLAM?
@akshat_sj VI
@akshat_sj which chip?
@akshat_sj visual slam on $10 is impressive but drifts fast without loops. what's your keyframe rejection strategy to keep the map manageable
How does it handle being picked up and set down somewhere else? That's the household test I'd want to see: a map it can recognise again after a human 'helps'.
@akshat_sj Could it run realtime on a phone?
@akshat_sj which slam framework is it optimiztaation based or filter based
![This seemingly obvious prediction didn't take long to become reality. MASt3R-SLAM: Real-Time Dense SLAM with 3D Reconstruction Priors Contributions: • The first real-time SLAM system using the two-view 3D reconstruction prior MASt3R [20] as a foundation. • Efficient techniques for pointmap matching, tracking and local fusion, graph construction and loop closure, and second-order global optimization. • A state-of-the-art dense SLAM system capable of handling generic, time-varying camera models. Abstract: We present a real-time monocular dense SLAM system, designed from the ground up using MASt3R, a two-view 3D reconstruction and matching prior. Equipped with this strong prior, our system remains robust on in-the-wild video sequences, making no assumptions on a fixed or parametric camera model beyond a unique camera center. Key features include: - Efficient methods for pointmap matching, camera tracking, and local fusion - Graph construction and loop closure - Second-order global optimization With known calibration, a simple modification achieves state-of-the-art performance across various benchmarks. Altogether, we propose a plug-and-play monocular SLAM system capable of producing globally-consistent poses and dense geometry while operating at 15 FPS.](https://image.24vids.com/tw-1868933842280604139/ext_tw_video_thumb/1868932917897871360/pu/img/WAmSelTOk2HJLm86.jpg)
