🥳MMDetection3D Release v1.2.0! - Support 3D Occupancy Prediction task and TPVFormer, a Camera-only LIDAR Semantic Segmentation model, and support LIDAR Semantic Segmentation on NuScenes Dataset. - Support BEVFusion, a LiDAR and Camera fusion model. #cv #AI

🔥Introducing the Largest 3D Occupancy Prediction Benchmark in autonomous driving: medium.com/@opendrivelab/… Check out our GitHub (github.com/OpenDriveLab/O…) for more details on the dataset, leaderboard, and upcoming challenge in 2024 (opendrivelab.com/AD24Challenge.…).

🌪️ Despite the ongoing super Typhoon Saola in Hong Kong, I'm excited to introduce PointLLM! 🌈🔍 It's a multi-modal large language model that understands point clouds. 1/4🧵 🔗 Demo: http://101.230.144.196 📄 Paper: arxiv.org/abs/2308.16911 💻 Code: github.com/OpenRobotLab/P…

🔥Unified Human-Scene Interaction via Prompted Chain-of-Contacts🔥 #UniHSI Code: github.com/OpenRobotLab/U… Project Page: xizaoqu.github.io/unihsi/ ArXiv: arxiv.org/abs/2309.07918 Hugging Face: huggingface.co/papers/2309.07…

Seeking a policy that can empower your robot to traverse any terrain? Our Hybrid Internal Model achieves this easily and only costs 1 hour. Key Insight: Estimating environmental dynamics with the robot's response. ArXiv: arxiv.org/abs/2312.11460 Code: github.com/OpenRobotLab/H…

Welcome to try the embodied AI track! We make a preliminary attempt with the multi-view 3D visual grounding benchmark. Cannot wait to see innovative submissions beating our baselines and top our benchmark! Thanks the great efforts from OpenDriveLab & other organizers!

Learning H-Infinity Locomotion Control Stable locomotion in precipitous environments is an essential capability of quadruped robots, demanding the ability to resist various external disturbances. However, recent learning-based policies only use basic domain randomization to

We are excited to introduce #GRUtopia! The first simulated **city-scale** interactive 3D society designed for various robots that serve humans! Integrate GRScenes, GRResidents, and GRBench! 100k+ Interactive scenes, 89 scene categories~ Code: github.com/OpenRobotLab/G…

LLaVA-3D A Simple yet Effective Pathway to Empowering LMMs with 3D-awareness Recent advancements in Large Multimodal Models (LMMs) have greatly enhanced their proficiency in 2D visual understanding tasks, enabling them to effectively process and understand images and videos.

Imagine a future where you can ask humanoid robots to clean your room, but some items, like heavy sofas, are too challenging for just one robot to move. Introducing CooHOI, a learning-based framework designed for the cooperative transportation of objects by multiple humanoid

How to achieve 3D perception without reconstructed point clouds or additional training, using only generalizable 2D and language foundation models? At #CoRL2024, we introduce VLM-Grounder, a zero-shot VLM agent for 3D visual grounding. Paper: arxiv.org/abs/2410.13860 with codes.

Excited to introduce the Perceptive Internal Model (PIM) for Humanoid Robots! The first policy simultaneously for: - Go up and down stairs, jump gaps, and 50cm high platforms. - Indoor and outdoor scenarios. - Unitree H1 and Fourier GR-1 robots. Paper: arxiv.org/abs/2411.14386

🫰Thrilled to introduce HOMIE: Humanoid Loco-Manipulation with Isomorphic Exoskeleton Cockpit. Website: homietele.github.io Code: github.com/OpenRobotLab/O… YouTube: youtu.be/FxkGmjyMc5g 😀 HOMIE consists of a novel RL-based training framework and a self-designed hardware

📢📢📢Excited to announce the 5th Workshop on 3D Scene Understanding for Vision, Graphics, and Robotics at #CVPR2025! Expect our awesome speakers and challenges on multi-modal 3D scene understanding and reasoning. 🎉🎉🎉#CVPR2025 Learn more at scene-understanding.com.

🤖Can we build a generalized robot navigation policy without any real-robot data? 👏We introduce the NavDP, which can zero-shot adapt to different robots in the open world. Website: wzcai99.github.io/navigation-dif… Github: github.com/wzcai99/NavDP/ Arxiv: arxiv.org/abs/2505.08712

StreamVLN Streaming Vision-and-Language Navigation via SlowFast Context Modeling