{"id":{"repo_id":"nus","oai_identifier":"oai:scholarbank.nus.edu.sg:10635/195563"},"canonical_url":"https://search.dev.ndltd.org/etd/nus/oai:scholarbank.nus.edu.sg:10635/195563","repository":{"repo_id":"nus","name":"National University of Singapore","base_url":"https://scholarbank.nus.edu.sg/oai/request"},"display":{"title":"VRAG: Region Attention Graph for Content-Based Video Retrieval","abstract":"Content-based Video Retrieval (CBVR) is used on media-sharing platforms for applications such as video recommendation and filtering. To manage databases that scale to billions of videos, video-level approaches that use fixed-size embeddings are preferred due to their efficiency. In this paper, we introduce Video Region Attention Graph Networks (VRAG) that improves the state-of-the-art of video-level methods. We represent videos at a finer granularity viaregion-level features and encode video spatio-temporal dy-namics through region-level relations. Our VRAG capturesthe relationships between regions based on their semantic content via self-attention and the permutation invariant aggregation of Graph Convolution. In addition, we show that the performance gap between video-level and frame-level methods can be reduced by segmenting videos into shots and using shot embeddings for video retrieval. We evaluate our VRAG over several video retrieval tasks and achieve a new state-of-the-art for video-level retrieval. Furthermore, `our shot-level VRAG shows higher retrieval precision than other existing video-level methods, and closer performance to frame-level methods at faster evaluation speeds. Finally, our code will be made publicly available.","abstract_html":"Content-based Video Retrieval (CBVR) is used on media-sharing platforms for applications such as video recommendation and filtering. To manage databases that scale to billions of videos, video-level approaches that use fixed-size embeddings are preferred due to their efficiency. In this paper, we introduce Video Region Attention Graph Networks (VRAG) that improves the state-of-the-art of video-level methods. We represent videos at a finer granularity viaregion-level features and encode video spatio-temporal dy-namics through region-level relations. Our VRAG capturesthe relationships between regions based on their semantic content via self-attention and the permutation invariant aggregation of Graph Convolution. In addition, we show that the performance gap between video-level and frame-level methods can be reduced by segmenting videos into shots and using shot embeddings for video retrieval. We evaluate our VRAG over several video retrieval tasks and achieve a new state-of-the-art for video-level retrieval. Furthermore, `our shot-level VRAG shows higher retrieval precision than other existing video-level methods, and closer performance to frame-level methods at faster evaluation speeds. Finally, our code will be made publicly available.","abstract_has_math":false,"creators":["KENNARD NG POOL HUA"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2021,"date_issued":"2021-04-22","date_published":"2021-04-22","updated_at":"2026-07-24T03:31:13Z","subjects":["Video retrieval"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["KENNARD NG POOL HUA"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2021-04-22"]},{"key":"dc:relation.isreferencedby","label":"Dc Relation Isreferencedby","values":["https://scholarbank.nus.edu.sg/handle/10635/195563"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Video retrieval"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://scholarbank.nus.edu.sg/bitstreams/b6c89303-22f5-41f7-b2ab-138eb6366b2d/download"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Content-based Video Retrieval (CBVR) is used on media-sharing platforms for applications such as video recommendation and filtering. To manage databases that scale to billions of videos, video-level approaches that use fixed-size embeddings are preferred due to their efficiency. In this paper, we introduce Video Region Attention Graph Networks (VRAG) that improves the state-of-the-art of video-level methods. We represent videos at a finer granularity viaregion-level features and encode video spatio-temporal dy-namics through region-level relations. Our VRAG capturesthe relationships between regions based on their semantic content via self-attention and the permutation invariant aggregation of Graph Convolution. In addition, we show that the performance gap between video-level and frame-level methods can be reduced by segmenting videos into shots and using shot embeddings for video retrieval. We evaluate our VRAG over several video retrieval tasks and achieve a new state-of-the-art for video-level retrieval. Furthermore, `our shot-level VRAG shows higher retrieval precision than other existing video-level methods, and closer performance to frame-level methods at faster evaluation speeds. Finally, our code will be made publicly available."]},{"key":"dc:format.checksum.md5","label":"Dc Format Checksum Md5","values":["da6c9cce5bf508c947044c29674f3bfe","d9e4cc5356d2762297f5edbc987c5021"]},{"key":"dc:title","label":"Title","values":["VRAG: Region Attention Graph for Content-Based Video Retrieval"]}]}],"canonical_facts":{"dc:creator":["KENNARD NG POOL HUA"],"dc:date.issued":["2021-04-22"],"dc:description.abstract":["Content-based Video Retrieval (CBVR) is used on media-sharing platforms for applications such as video recommendation and filtering. To manage databases that scale to billions of videos, video-level approaches that use fixed-size embeddings are preferred due to their efficiency. In this paper, we introduce Video Region Attention Graph Networks (VRAG) that improves the state-of-the-art of video-level methods. We represent videos at a finer granularity viaregion-level features and encode video spatio-temporal dy-namics through region-level relations. Our VRAG capturesthe relationships between regions based on their semantic content via self-attention and the permutation invariant aggregation of Graph Convolution. In addition, we show that the performance gap between video-level and frame-level methods can be reduced by segmenting videos into shots and using shot embeddings for video retrieval. We evaluate our VRAG over several video retrieval tasks and achieve a new state-of-the-art for video-level retrieval. Furthermore, `our shot-level VRAG shows higher retrieval precision than other existing video-level methods, and closer performance to frame-level methods at faster evaluation speeds. Finally, our code will be made publicly available."],"dc:format.checksum.md5":["da6c9cce5bf508c947044c29674f3bfe","d9e4cc5356d2762297f5edbc987c5021"],"dc:identifier.uri":["https://scholarbank.nus.edu.sg/bitstreams/b6c89303-22f5-41f7-b2ab-138eb6366b2d/download"],"dc:relation.isreferencedby":["https://scholarbank.nus.edu.sg/handle/10635/195563"],"dc:subject":["Video retrieval"],"dc:title":["VRAG: Region Attention Graph for Content-Based Video Retrieval"],"dc:type":["Thesis"]},"updated_at":"2026-07-24T03:31:13Z"}