Obj-GloVe: Scene-Based Contextual Object Embedding.
Canwen XuZhenzhong ChenChenliang LiPublished in: CoRR (2019)
Keyphrases
- multiple objects
- visual context
- complex scenes
- moving objects
- spatial relations
- real world scenes
- object models
- real world objects
- target object
- object features
- reference object
- ground plane
- d objects
- video sequences
- geometric information
- relative position
- individual objects
- acquired images
- background pixels
- single image
- three dimensional
- location and orientation
- wire frame
- contextual cues
- uncalibrated images
- image sequences
- visual scene
- real objects
- intensity images
- viewing angle
- video scene
- object model
- contextual information
- object appearance
- cast shadows
- scene interpretation
- image segments
- lighting conditions
- image regions
- human body
- viewing position