Shape-Former: Bridging CNN and Transformer via ShapeConv for multimodal image matching.
Jiaxuan ChenXiaoxian ChenShuang ChenYuyan LiuYujing RaoYang YangHaifeng WangDan WuPublished in: Inf. Fusion (2023)
Keyphrases
- image matching
- shape matching
- invariant features
- image pairs
- keypoints
- object recognition
- affine transformation
- matching algorithm
- computer vision
- feature matching
- cellular neural networks
- template matching
- object identification
- shape features
- image matching method
- image transformations
- shape analysis
- shape descriptors
- image retrieval
- wide baseline matching
- illumination variations
- stereo matching
- image registration
- feature descriptors
- image set
- matching scheme
- wide baseline
- affine distortion
- sift descriptors
- machine learning
- multiscale
- affine invariant
- pattern recognition
- viewpoint