A professional technology event features a demonstration area showcasing a presentation titled "SceneFun3D: Fine-grained Functionality and Affordance Understanding in 3D Scenes." The presentation covers various aspects of 3D scene understanding, including Functionality Annotations, Natural Language Task Descriptions, and Motion Annotations, displayed on a large screen mounted on a modern, slatted wooden wall. A person stands beside a laptop on a small wooden table, engaging with the content on the screen, which mirrors the theme of the presentation. The surroundings feature a bustling event atmosphere with other attendees and exhibitors in the background, adding to the vibrant and innovative ambiance of the event.
Text transcribed from the image:
Creativity Enrich Life
Inspire Creativity Enry
4444444444
3D Scene Understanding
Open-Vocabulary with Foundation Models
Functionalities & Affordances
SceneFun3D: Fine-grained Functionality and Affordance Understanding in 3D Scenes
Functionality
Annotations
Natural Language
Task Descriptions
hook_turn
hook_pull
hook_turn
key_press
pinch_pull
foot_push
hook_pull
plug_in
tip_push
unplug
Open the oven door
ender
[1] Delitzas et al. "SceneFun3D: Fine-grained Functionality and Affordance Understanding in 3D Scenes CVPR 24
ETH zürich Francis Engelmann
36
2nd Challenge on Open Vocabulary 3D Scene Understanding
Tack 1:30 Object Instance Search
Motion
Annotations
DINEK