Been playing with #SpatialLM - an LLM for 3D PCD. It turns scans/video → walls/doors/windows + oriented 3D boxes. Where it actually feels useful → quick indoor mapping for robotics, BIMs, claims/inspections. From the paper, the outputs are editable, so anyone can correct a

SpatialLM just dropped on Hugging Face Large Language Model for Spatial Understanding

SpatialLM is wild This AI turns simple videos into accurate 3D layouts & floorplans instantly. How it works, and how to access👇

SpatialLM is trained on realistic 3D data to accurately match real world rooms and objects.

SpatialLM is super cool, but it does NOT work in real time. To make 3D floor plans on the fly use Apple’s Room Plan API — available in the Polycam app. Computer vision that is truly magical — and it does everything locally on your LiDAR enabled iPhone! https://t.co/oot1aRyjZu

یک مدل زبان بزرگ سه‌بعدی (3D LLM) به اسم SpatialLM اومده که برای پردازش داده‌های 3D point cloud طراحی شده و خروجی‌های ساختار یافته از درک صحنه‌های 3D تولید می‌کند. این خروجی‌ها شامل عناصری معماری مانند دیوارها، درها، پنجره‌ها همراه با دسته‌بندی‌های معنایی‌شان هستند. این مدل

Funtime with SpatialLM- eventually it will serve well in embodied AI.

SpatialLM is very interesting it's a new model that encodes text prompts (e.g. detect windows) and point clouds projected to an LLM 🤯 the LLM outputs 3D bounding boxes, very simple yet effective approach two models based on Llama and Qwen-0.5B and Llama-1B on @huggingface

还记得3月份给大家介绍的 SpatialLM 空间推理模型吗? 他们刚刚更新了 SpatialLM-1.1 版本! 这次的新版本特性包括: - 支持多种输入源模式： - 基于AI的文本生成3D场景（Text-to-3D， 大模型生成的图像转3D都行) - 手持摄像机视频重建 - LiDAR点云数据 (iPhone Pro的激光雷达就可以) -

SpatialLM : 3D large language model designed to process 3D point cloud data and generate structured 3D scene understanding outputs. Available on Huggingface 👇

SpatialLM，这个 3D 视觉大模型，感觉可以被改造成实时场景背词神器，1 秒生成英文词汇墙🤪 摄像头扫过之处，所有家具都会自动生成英文标签： 餐桌弹出“dinner table”，沙发标注“sofa”... 这种所见即所得的认知轰炸，分分钟词汇量暴涨啊 链接见评论👇

什么是空间大模型？ 帮大家梳理了上周末的群核 TechDay 的内容, 包括 SpatialLM 和 SpatialGen 两个空间模型各自的能力和应用场景. 群核提出的空间大模型概念非常有意思, 这要求大模型不但能感知空间, 还可以生成空间的模型. 从而实现大模型对空间能力的"自举" 希望对大家有帮助~ #群核科技

SpatialLM. Fast, not real-time. So much to explore here. I see the potential. More coming soon. #AI #AEC

VLM and SpatialLM robotics are the future.

スマホカメラから実世界の3D空間を理解するLLM「SpatialLM」、静止画像を3Dビデオに変換する「Stable Virtual Camera」など生成AI技術5つを解説（生成AIウィークリー）  https://t.co/lAtBWjkhsq 単眼カメラに映るソファやドアなどをタグ名付きのバウンディングボックスで出力するLLM「SpatialLM」

By combining CNN Open Vocabulary Detection, Visual Language Models, and the powerful SpatialLM applied to 3D point clouds, we're about to transform the future of artificial intelligence applied to facial identification. Get ready for innovations that go beyond what meets the eye.

SpatialLM turns any 3D point cloud—from phone video, RGBD, or LiDAR—into structured scene understanding with walls, doors, and object boxes. No special equipment needed for what used to require expensive sensors.

Meski ini terjadi secara post-processing, tetep aja bakal berguna untuk membantu orang buta mengenali sekelilingnya. SpatialLM: Large Language Model for Spatial Understanding https://t.co/DSwDURnqQR

This model is soo cool! If you are able to fine tune similar models further, you can build a much dynamic assembly line with robots. Model: SpatialLM

SpatialLM just dropped on Hugging Face Large Language Model for Spatial Understanding Model 👇