{"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/multimodal-representation-learning-using-deep","title":"Multimodal Representation Learning using Deep Multiset Canonical Correlation","arxiv_id":"1904.01775","date":"2019-04-03","proceeding":null,"authors":["Krishna Somandepalli","Naveen Kumar","Ruchir Travadi","Shrikanth Narayanan"],"abstract":"We propose Deep Multiset Canonical Correlation Analysis (dMCCA) as an\nextension to representation learning using CCA when the underlying signal is\nobserved across multiple (more than two) modalities. We use deep learning\nframework to learn non-linear transformations from different modalities to a\nshared subspace such that the representations maximize the ratio of between-\nand within-modality covariance of the observations. Unlike linear discriminant\nanalysis, we do not need class information to learn these representations, and\nwe show that this model can be trained for complex data using mini-batches.\nUsing synthetic data experiments, we show that dMCCA can effectively recover\nthe common signal across the different modalities corrupted by multiplicative\nand additive noise. We also analyze the sensitivity of our model to recover the\ncorrelated components with respect to mini-batch size and dimension of the\nembeddings. Performance evaluation on noisy handwritten datasets shows that our\nmodel outperforms other CCA-based approaches and is comparable to deep neural\nnetwork models trained end-to-end on this dataset.","url_abs":"http://arxiv.org/abs/1904.01775v1","url_pdf":"http://arxiv.org/pdf/1904.01775v1.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":[{"paper_slug":"multimodal-representation-learning-using-deep","repo_url":"https://github.com/usc-sail/mica-deep-mcca","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}