Sketch Me if You Can: Towards Generating Detailed Descriptions of Object Shape by Grounding in Images and Drawings.
Ting HanSina ZarrießPublished in: INLG (2019)
Keyphrases
- object shape
- detailed descriptions
- object recognition
- image analysis
- image data
- object boundaries
- image features
- object localization
- image classification
- input image
- object surface
- shape representation
- image retrieval
- three dimensional
- computer vision
- image registration
- illumination conditions
- object models
- keypoints
- image regions
- image matching
- region of interest
- object segmentation
- bounding box
- shape variations
- appearance model
- shape descriptors
- lighting conditions
- segmentation method
- case study