{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/88970"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/88970","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Transform learning based image and video processing","abstract":"In recent years, sparse signal modeling, especially using the synthesis dictionary model, has received much attention. Sparse coding in the synthesis model is, however, NP-hard. Various methods have been proposed to learn such synthesis dictionaries from data. Numerous applications such as image denoising, magnetic resonance image (MRI), and computed tomography (CT) reconstruction have been shown to benefit from a good adaptive sparse model. Recently, the sparsifying transform model has received interest, for which sparse coding is cheap and exact, and learning, or data-driven adaptation admits computationally efficient solutions. In this thesis, we present two extensions to the transform learning framework, and some applications. In the first part of this thesis, we propose a union of sparsifying transforms model. Sparse coding in this model reduces to a form of clustering. The proposed model is also equivalent to a structured overcomplete sparsifying transform model with block cosparsity, dubbed OCTOBOS. The alternating algorithm introduced for learning such transforms involves simple closed-form solutions. Theoretical analysis provides a convergence guarantee for this algorithm. It is shown to be globally convergent to the set of partial minimizers of the non-convex learning problem. When applied to images, the algorithm learns a collection of well-conditioned square transforms, and a good clustering of patches or textures. The resulting sparse representations for the images are better than those obtained with a single learned transform, or with analytical transforms. We show the promising performance of the proposed approach in image denoising, which compares quite favorably with approaches involving a single learned square transform or an overcomplete synthesis dictionary, or Gaussian mixture models. The proposed denoising method is also faster than the synthesis dictionary based approach. Next, we develop a methodology for online learning of square sparsifying transforms. Such online learning can be particularly useful when dealing with big data, and for signal processing applications such as real-time sparse representation and denoising. The proposed transform learning algorithms are shown to have a significantly lower computational cost than online synthesis dictionary learning. In practice, the sequential learning of a sparsifying transform typically converges faster than batch mode transform learning. Preliminary experiments show the usefulness of the proposed schemes for sparse representation, and denoising. In the third part, we present a video denoising framework based on online 3D sparsifying transform learning. The proposed scheme has low computational and memory costs, and can handle streaming video. Our numerical experiments show promising performance for the proposed video denoising method compared to popular prior or state-of-the-art methods.","abstract_html":"In recent years, sparse signal modeling, especially using the synthesis dictionary model, has received much attention. Sparse coding in the synthesis model is, however, NP-hard. Various methods have been proposed to learn such synthesis dictionaries from data. Numerous applications such as image denoising, magnetic resonance image (MRI), and computed tomography (CT) reconstruction have been shown to benefit from a good adaptive sparse model. Recently, the sparsifying transform model has received interest, for which sparse coding is cheap and exact, and learning, or data-driven adaptation admits computationally efficient solutions. In this thesis, we present two extensions to the transform learning framework, and some applications. In the first part of this thesis, we propose a union of sparsifying transforms model. Sparse coding in this model reduces to a form of clustering. The proposed model is also equivalent to a structured overcomplete sparsifying transform model with block cosparsity, dubbed OCTOBOS. The alternating algorithm introduced for learning such transforms involves simple closed-form solutions. Theoretical analysis provides a convergence guarantee for this algorithm. It is shown to be globally convergent to the set of partial minimizers of the non-convex learning problem. When applied to images, the algorithm learns a collection of well-conditioned square transforms, and a good clustering of patches or textures. The resulting sparse representations for the images are better than those obtained with a single learned transform, or with analytical transforms. We show the promising performance of the proposed approach in image denoising, which compares quite favorably with approaches involving a single learned square transform or an overcomplete synthesis dictionary, or Gaussian mixture models. The proposed denoising method is also faster than the synthesis dictionary based approach. Next, we develop a methodology for online learning of square sparsifying transforms. Such online learning can be particularly useful when dealing with big data, and for signal processing applications such as real-time sparse representation and denoising. The proposed transform learning algorithms are shown to have a significantly lower computational cost than online synthesis dictionary learning. In practice, the sequential learning of a sparsifying transform typically converges faster than batch mode transform learning. Preliminary experiments show the usefulness of the proposed schemes for sparse representation, and denoising. In the third part, we present a video denoising framework based on online 3D sparsifying transform learning. The proposed scheme has low computational and memory costs, and can handle streaming video. Our numerical experiments show promising performance for the proposed video denoising method compared to popular prior or state-of-the-art methods.","abstract_has_math":false,"creators":["Wen, Bihan"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Electrical & Computer Engineering","degree_department":null,"school":null,"contributors":["Bresler, Yoram"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2016,"date_issued":"2016-03-02T19:33:27Z","date_published":"2016-03-02T19:33:27Z","updated_at":"2026-07-22T22:26:32Z","subjects":["Dictionary Learning","Transform Learning","Sparse Representation","Online Learning","Image Representation","Image Denoising","Video Denoising","Image Segmentation","Clustering","Machine Learning","Overcomplete Representation","Big Data","Inverse Problem"],"languages":["en"],"rights":["Copyright 2015 Bihan Wen"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/88970","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Bresler, Yoram"]},{"key":"dc:creator","label":"Author","values":["Wen, Bihan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2016-03-02T19:33:27Z","2015-10-30","2015-12"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical & Computer Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Dictionary Learning","Transform Learning","Sparse Representation","Online Learning","Image Representation","Image Denoising","Video Denoising","Image Segmentation","Clustering","Machine Learning","Overcomplete Representation","Big Data","Inverse Problem"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2015 Bihan Wen"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/88970"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["In recent years, sparse signal modeling, especially using the synthesis dictionary model, has received much attention. Sparse coding in the synthesis model is, however, NP-hard. Various methods have been proposed to learn such synthesis dictionaries from data. Numerous applications such as image denoising, magnetic resonance image (MRI), and computed tomography (CT) reconstruction have been shown to benefit from a good adaptive sparse model. Recently, the sparsifying transform model has received interest, for which sparse coding is cheap and exact, and learning, or data-driven adaptation admits computationally efficient solutions. In this thesis, we present two extensions to the transform learning framework, and some applications. In the first part of this thesis, we propose a union of sparsifying transforms model. Sparse coding in this model reduces to a form of clustering. The proposed model is also equivalent to a structured overcomplete sparsifying transform model with block cosparsity, dubbed OCTOBOS. The alternating algorithm introduced for learning such transforms involves simple closed-form solutions. Theoretical analysis provides a convergence guarantee for this algorithm. It is shown to be globally convergent to the set of partial minimizers of the non-convex learning problem. When applied to images, the algorithm learns a collection of well-conditioned square transforms, and a good clustering of patches or textures. The resulting sparse representations for the images are better than those obtained with a single learned transform, or with analytical transforms. We show the promising performance of the proposed approach in image denoising, which compares quite favorably with approaches involving a single learned square transform or an overcomplete synthesis dictionary, or Gaussian mixture models. The proposed denoising method is also faster than the synthesis dictionary based approach. Next, we develop a methodology for online learning of square sparsifying transforms. Such online learning can be particularly useful when dealing with big data, and for signal processing applications such as real-time sparse representation and denoising. The proposed transform learning algorithms are shown to have a significantly lower computational cost than online synthesis dictionary learning. In practice, the sequential learning of a sparsifying transform typically converges faster than batch mode transform learning. Preliminary experiments show the usefulness of the proposed schemes for sparse representation, and denoising. In the third part, we present a video denoising framework based on online 3D sparsifying transform learning. The proposed scheme has low computational and memory costs, and can handle streaming video. Our numerical experiments show promising performance for the proposed video denoising method compared to popular prior or state-of-the-art methods.","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2016-03-02 without embargo terms","The student, Bihan Wen, accepted the attached license on 2015-10-22 at 11:26.","The student, Bihan Wen, submitted this Thesis for approval on 2015-10-28 at 15:47.","This Thesis was approved for publication on 2015-10-30 at 13:32.","DSpace SAF Submission Ingestion Package generated from Vireo submission #8736 on 2016-03-02 at 12:49:44","Made available in DSpace on 2016-03-02T19:33:27Z (GMT). No. of bitstreams: 2 WEN-THESIS-2015.pdf: 9119458 bytes, checksum: 19124c38290ee923a1123dbbcd1ed5fa (MD5) LICENSE.txt: 4206 bytes, checksum: 1120edf0b55b76abacb88f504b5aa546 (MD5) Previous issue date: 2015-10-30"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Transform learning based image and video processing"]}]}],"canonical_facts":{"dc:contributor":["Bresler, Yoram"],"dc:creator":["Wen, Bihan"],"dc:date":["2016-03-02T19:33:27Z","2015-10-30","2015-12"],"dc:description":["In recent years, sparse signal modeling, especially using the synthesis dictionary model, has received much attention. Sparse coding in the synthesis model is, however, NP-hard. Various methods have been proposed to learn such synthesis dictionaries from data. Numerous applications such as image denoising, magnetic resonance image (MRI), and computed tomography (CT) reconstruction have been shown to benefit from a good adaptive sparse model. Recently, the sparsifying transform model has received interest, for which sparse coding is cheap and exact, and learning, or data-driven adaptation admits computationally efficient solutions. In this thesis, we present two extensions to the transform learning framework, and some applications. In the first part of this thesis, we propose a union of sparsifying transforms model. Sparse coding in this model reduces to a form of clustering. The proposed model is also equivalent to a structured overcomplete sparsifying transform model with block cosparsity, dubbed OCTOBOS. The alternating algorithm introduced for learning such transforms involves simple closed-form solutions. Theoretical analysis provides a convergence guarantee for this algorithm. It is shown to be globally convergent to the set of partial minimizers of the non-convex learning problem. When applied to images, the algorithm learns a collection of well-conditioned square transforms, and a good clustering of patches or textures. The resulting sparse representations for the images are better than those obtained with a single learned transform, or with analytical transforms. We show the promising performance of the proposed approach in image denoising, which compares quite favorably with approaches involving a single learned square transform or an overcomplete synthesis dictionary, or Gaussian mixture models. The proposed denoising method is also faster than the synthesis dictionary based approach. Next, we develop a methodology for online learning of square sparsifying transforms. Such online learning can be particularly useful when dealing with big data, and for signal processing applications such as real-time sparse representation and denoising. The proposed transform learning algorithms are shown to have a significantly lower computational cost than online synthesis dictionary learning. In practice, the sequential learning of a sparsifying transform typically converges faster than batch mode transform learning. Preliminary experiments show the usefulness of the proposed schemes for sparse representation, and denoising. In the third part, we present a video denoising framework based on online 3D sparsifying transform learning. The proposed scheme has low computational and memory costs, and can handle streaming video. Our numerical experiments show promising performance for the proposed video denoising method compared to popular prior or state-of-the-art methods.","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2016-03-02 without embargo terms","The student, Bihan Wen, accepted the attached license on 2015-10-22 at 11:26.","The student, Bihan Wen, submitted this Thesis for approval on 2015-10-28 at 15:47.","This Thesis was approved for publication on 2015-10-30 at 13:32.","DSpace SAF Submission Ingestion Package generated from Vireo submission #8736 on 2016-03-02 at 12:49:44","Made available in DSpace on 2016-03-02T19:33:27Z (GMT). No. of bitstreams: 2 WEN-THESIS-2015.pdf: 9119458 bytes, checksum: 19124c38290ee923a1123dbbcd1ed5fa (MD5) LICENSE.txt: 4206 bytes, checksum: 1120edf0b55b76abacb88f504b5aa546 (MD5) Previous issue date: 2015-10-30"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/88970"],"dc:language":["en"],"dc:rights":["Copyright 2015 Bihan Wen"],"dc:subject":["Dictionary Learning","Transform Learning","Sparse Representation","Online Learning","Image Representation","Image Denoising","Video Denoising","Image Segmentation","Clustering","Machine Learning","Overcomplete Representation","Big Data","Inverse Problem"],"dc:title":["Transform learning based image and video processing"],"dc:type":["text"],"thesis:degree_discipline":["Electrical & Computer Engineering"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:26:32Z"}