ECCV Demo:
Language-guided Scene Interaction

Thanks for checking our demo! Below we list all details and individual works that have been combined into this demo or sparked further research.

SpotLight Logo

A Framework for Open-Vocabulary Object Retrieval and
Drawer Manipulation in Point Clouds

1ETH Zürich 2Microsoft 3Google
SpotLight Logo

Robotic Scene Understanding through Interaction and Affordance Detection

1ETH Zürich, 2Microsoft 3Uni Bonn

Related and Inspired Works

OpenMask3D

Open-Vocabulary 3D Instance Segmentation

LabelMaker

Accurate 2D and 3D Semantic Labels from RGB-D Scans

OpenDAS

Domain Adaptation for Open-Vocabulary Segmentation

ICGNet

A Unified Approach for Instance-Centric Grasping

HoloLens-ROS

Operating Robots through Mixed Reality