Dataset Groups Activity Stream AVSBench Audio-visual segmentation (AVS) aims to segment sound sources in the video sequence, requiring a pixel-level understanding of audio-visual correspondence. BibTex: @dataset{Juhyeong_Seon_and_Woobin_Im_and_Sebin_Lee_and_Jumin_Lee_and_Sung-Eui_Yoon_2024, abstract = {Audio-visual segmentation (AVS) aims to segment sound sources in the video sequence, requiring a pixel-level understanding of audio-visual correspondence.}, author = {Juhyeong Seon and Woobin Im and Sebin Lee and Jumin Lee and Sung-Eui Yoon}, doi = {10.57702/n0dvi342}, institution = {No Organization}, keyword = {'5-second videos', 'AVSBench', 'Audio-Visual Segmentation', 'Dataset', 'Image Analysis', 'Object Detection', 'audio-visual segmentation', 'pixel-wise annotations', 'sound source localization', 'video analysis'}, month = {dec}, publisher = {TIB}, title = {AVSBench}, url = {https://service.tib.eu/ldmservice/dataset/avsbench}, year = {2024} }
