Matching Non-Identical Objects
This addresses a novel task for computer vision applications involving diverse object variations, but it appears incremental as it builds on existing sparse matching methods.
The study tackled the problem of matching non-identical but similar objects at the pixel level by incorporating semantic information from object detectors into sparse image matching methods, achieving successful matching across cases like in-class variations, class discrepancies, and domain shifts.
Not identical but similar objects are ubiquitous in our world, ranging from four-legged animals such as dogs and cats to cars of different models and flowers of various colors. This study addresses a novel task of matching such non-identical objects at the pixel level. We propose a weighting scheme of descriptors that incorporates semantic information from object detectors into existing sparse image matching methods, extending their targets from identical objects captured from different perspectives to semantically similar objects. The experiments show successful matching between non-identical objects in various cases, including in-class design variations, class discrepancy, and domain shifts (e.g., photo--drawing and image corruptions).