OCID-Ref: A 3D Robotic Dataset With Embodied Language For Clutter Scene Grounding.
Ke-Jyun WangYun-Hsuan LiuHung-Ting SuJen-Wei WangYu-Siang WangWinston H. HsuWen-Chin ChenPublished in: NAACL-HLT (2021)
Keyphrases
- indoor and outdoor scenes
- d scene
- three dimensional
- programming language
- outdoor images
- image sequences
- video sequences
- single image
- language learning
- complex scenes
- mobile robot
- multiple images
- natural language
- benchmark datasets
- scene analysis
- scene understanding
- robotic systems
- scene classification
- image transformations
- cognitive process
- manipulation tasks
- real time
- cognitive science
- dynamic scenes
- moving objects
- image set
- intelligent agents
- input image
- object oriented
- optical flow