{"scopus_import":"1","conference":{"location":"Bellevue, United States","start_date":"2011-06-28","name":"ICML: International Conference on Machine Learning","end_date":"2011-07-02"},"citation":{"chicago":"Quadrianto, Novi, and Christoph Lampert. “Learning Multi-View Neighborhood Preserving Projections,” 425–32. ML Research Press, 2011.","ama":"Quadrianto N, Lampert C. Learning multi-view neighborhood preserving projections. In: ML Research Press; 2011:425-432.","ieee":"N. Quadrianto and C. Lampert, “Learning multi-view neighborhood preserving projections,” presented at the ICML: International Conference on Machine Learning, Bellevue, United States, 2011, pp. 425–432.","ista":"Quadrianto N, Lampert C. 2011. Learning multi-view neighborhood preserving projections. ICML: International Conference on Machine Learning, 425–432.","mla":"Quadrianto, Novi, and Christoph Lampert. <i>Learning Multi-View Neighborhood Preserving Projections</i>. ML Research Press, 2011, pp. 425–32.","apa":"Quadrianto, N., &#38; Lampert, C. (2011). Learning multi-view neighborhood preserving projections (pp. 425–432). Presented at the ICML: International Conference on Machine Learning, Bellevue, United States: ML Research Press.","short":"N. Quadrianto, C. Lampert, in:, ML Research Press, 2011, pp. 425–432."},"publication_status":"published","oa_version":"None","article_processing_charge":"No","language":[{"iso":"eng"}],"abstract":[{"lang":"eng","text":"We address the problem of metric learning for multi-view data, namely the construction of embedding projections from data in different representations into a shared feature space, such that the Euclidean distance in this space provides a meaningful within-view as well as between-view similarity. Our motivation stems from the problem of cross-media retrieval tasks, where the availability of a joint Euclidean distance function is a pre-requisite to allow fast, in particular hashing-based, nearest neighbor queries. We formulate an objective function that expresses the intuitive concept that matching samples are mapped closely together in the output space, whereas non-matching samples are pushed apart, no matter in which view they are available. The resulting optimization problem is not convex, but it can be decomposed explicitly into a convex and a concave part, thereby allowing efficient optimization using the convex-concave procedure. Experiments on an image retrieval task show that nearest-neighbor based cross-view retrieval is indeed possible, and the proposed technique improves the retrieval accuracy over baseline techniques."}],"publisher":"ML Research Press","publist_id":"3316","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","department":[{"_id":"ChLa"}],"_id":"3319","title":"Learning multi-view neighborhood preserving projections","page":"425 - 432","date_updated":"2023-10-17T11:59:50Z","month":"01","date_created":"2018-12-11T12:02:39Z","type":"conference","day":"01","year":"2011","status":"public","author":[{"first_name":"Novi","last_name":"Quadrianto","full_name":"Quadrianto, Novi"},{"orcid":"0000-0001-8622-7887","last_name":"Lampert","full_name":"Lampert, Christoph","first_name":"Christoph","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87"}],"date_published":"2011-01-01T00:00:00Z"}