InternRobotics/VLM-Grounder
[CoRL 2024] VLM-Grounder: A VLM Agent for Zero-Shot 3D Visual Grounding
This project helps robotics engineers and researchers precisely identify and locate specific objects within a 3D environment using only 2D camera images. You input a sequence of 2D images and a text description of a target object, and the system outputs the estimated 3D bounding box for that object. It's designed for those working with autonomous systems, robotic manipulation, or augmented reality applications who need to 'ground' linguistic descriptions to physical objects without relying on pre-existing 3D models or object knowledge.
129 stars. No commits in the last 6 months.
Use this if you need to pinpoint an object's exact 3D location in a room or scene, based on a simple text description and a sequence of camera images, especially when a 3D model of the scene or object isn't available.
Not ideal if your application requires real-time object grounding on a single image without any sequential context, or if you primarily work with pre-scanned 3D point clouds rather than camera feeds.
Stars
129
Forks
2
Language
Python
License
—
Category
Last pushed
May 22, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/InternRobotics/VLM-Grounder"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
jingyaogong/minimind-v
🚀 「大模型」1小时从0训练26M参数的视觉多模态VLM！🌏 Train a 26M-parameter VLM from scratch in just 1 hours!
SkyworkAI/Skywork-R1V
Skywork-R1V is an advanced multimodal AI model series developed by Skywork AI, specializing in...
roboflow/vision-ai-checkup
Take your LLM to the optometrist.
zai-org/GLM-TTS
GLM-TTS: Controllable & Emotion-Expressive Zero-shot TTS with Multi-Reward Reinforcement Learning
NExT-GPT/NExT-GPT
Code and models for ICML 2024 paper, NExT-GPT: Any-to-Any Multimodal Large Language Model