{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deep-spectral-correspondence-for-matching","title":"Deep Spectral Correspondence for Matching Disparate Image Pairs","arxiv_id":"1809.04642","date":"2018-09-12","proceeding":null,"authors":["Arun CS Kumar","Shefali Srivastava","Anirban Mukhopadhyay","Suchendra M. Bhandarkar"],"abstract":"A novel, non-learning-based, saliency-aware, shape-cognizant correspondence\ndetermination technique is proposed for matching image pairs that are\nsignificantly disparate in nature. Images in the real world often exhibit high\ndegrees of variation in scale, orientation, viewpoint, illumination and affine\nprojection parameters, and are often accompanied by the presence of textureless\nregions and complete or partial occlusion of scene objects. The above\nconditions confound most correspondence determination techniques by rendering\nimpractical the use of global contour-based descriptors or local pixel-level\nfeatures for establishing correspondence. The proposed deep spectral\ncorrespondence (DSC) determination scheme harnesses the representational power\nof local feature descriptors to derive a complex high-level global shape\nrepresentation for matching disparate images. The proposed scheme reasons about\ncorrespondence between disparate images using high-level global shape cues\nderived from low-level local feature descriptors. Consequently, the proposed\nscheme enjoys the best of both worlds, i.e., a high degree of invariance to\naffine parameters such as scale, orientation, viewpoint, illumination afforded\nby the global shape cues and robustness to occlusion provided by the low-level\nfeature descriptors. While the shape-based component within the proposed scheme\ninfers what to look for, an additional saliency-based component dictates where\nto look at thereby tackling the noisy correspondences arising from the presence\nof textureless regions and complex backgrounds. In the proposed scheme, a joint\nimage graph is constructed using distances computed between interest points in\nthe appearance (i.e., image) space. Eigenspectral decomposition of the joint\nimage graph allows for reasoning about shape similarity to be performed\njointly, in the appearance space and eigenspace.","url_abs":"http://arxiv.org/abs/1809.04642v1","url_pdf":"http://arxiv.org/pdf/1809.04642v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"matching-disparate-images","task_name":"Matching Disparate Images"}],"methods":[],"datasets_introduced":[{"slug":"dispscenes","name":"DispScenes","full_name":null}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}