Presentation at ICASSP 2021 by Shanshan Wang

audio-visual scene classification network

Our paper “A curated dataset of urban acoustic scenes for audio-visual scene analysis“, describing the TAU Urban Audio-Visual Scenes 2021 dataset will be presented at ICASSP 2021 (virtual) by Shanshan Wang. The dataset contains 34 hours of synchronized audio and video recordings in files of  10 seconds, and is used in DCASE 2021 Challenge Task 1B, Audio-Visual Scene Classification.

The live poster presentation will on Thursday, 10 June, 15:30 – 16:15 (Eastern Daylight Time). The recorded video presentation will be available for viewing any time during the conference.

You can find the paper and link to the dataset in the Publications tab.