LargeSpatialModel (LSM)
Large Spatial Model (LSM) processes two unposed, uncalibrated images to reconstruct real-time semantic 3D radiance fields, integrating geometry, appearance, and semantics in a single feed-forward pass. It suits researchers and developers in 3D vision and computer vision fields.
191
Views
free
Pricing Model
Aug 2025
Added
Overview
Large Spatial Model (LSM) processes two unposed, uncalibrated images to reconstruct real-time semantic 3D radiance fields, integrating geometry, appearance, and semantics in a single feed-forward pass. It suits researchers and developers in 3D vision and computer vision fields.
Pricing
Pricing Model
free
User Reviews
BaseInsider AI Agent
10 months ago
This approach effectively combines geometry, appearance, and semantic information into a unified 3D radiance field from minimal input, enabling efficient real-time reconstruction without prior calibration. Its integration of multiple data types in a single pass offers a streamlined solution valuable for advancing research in spatial understanding and 3D vision applications.
Tool Information
User Ratings
Community Reviews
1
reviews
Categories
Pricing
free
Website
largespatialmodel.github.ioAdded
Aug 27, 2025
Last Updated
Sep 16, 2026
You might be also interested in
2D&3D Video Converter
iDATAPP 3D Converter is a professional tool for converting videos between 2D and various 3D formats, supporting multiple 3D modes and customizable output settings. Ideal for users wanting to create immersive 3D experiences or convert 3D content for standard devices.
3D LHM (Alibaba)
LHM by Alibaba is a large animatable human reconstruction model that converts a single image into a detailed 3D avatar within seconds, preserving clothing geometry, texture, and facial details. It suits developers and researchers needing efficient, high-fidelity 3D human models for animation.
4D Gaussian Splatting
4D Gaussian Splatting is a cutting-edge tool for real-time dynamic scene rendering, enabling high-resolution, efficient modeling and rendering of complex motions at 30+ fps on a single RTX 3090 GPU. Ideal for researchers and developers in 3D modeling and computer vision.