Object2Scene: Putting Objects in Context for Open-Vocabulary 3D Detection.
Chenming ZhuWenwei ZhangTai WangXihui LiuKai ChenPublished in: CoRR (2023)
Keyphrases
- multiple objects
- visual scene
- cluttered scenes
- complex scenes
- object models
- real world scenes
- moving objects
- detecting objects
- individual objects
- real objects
- object detectors
- target object
- visual context
- video scene
- real world objects
- object detection
- d objects
- object model
- spatial relations
- bounding box
- object features
- parking lot
- object detection and tracking
- object geometry
- contextual cues
- man made
- object classes
- crowded scenes
- partial occlusion
- relative position
- physical objects
- data objects
- uncalibrated images
- reference object
- cluttered background
- object appearance
- tracked objects
- object segmentation
- similar objects
- object trajectories
- wire frame
- visual input
- image segments
- object hypotheses
- video sequences
- object parts
- cast shadows
- object tracking
- dynamic scenes
- geometric models
- spatial location
- object recognition
- geometric information
- background clutter
- object motion
- scene understanding
- image regions
- background pixels
- image sequences
- particle filter
- video data
- spatial relationships
- multiple images
- curved surfaces
- viewing angle
- lighting conditions
- spatial configuration
- d scene
- object instances
- ground plane
- rigid transformation
- image features