How to Design a Three-Stage Architecture for Audio-Visual Active Speaker Detection in the Wild
This work addresses the challenge of accurately identifying active speakers in real-world audio-visual scenarios, which is incremental as it builds on existing methods with specific architectural improvements.
The paper tackles the problem of active speaker detection in audio-visual data by proposing a three-stage pipeline and a new architecture called ASDNet, which achieves a state-of-the-art mAP of 93.5% on the AVA-ActiveSpeaker dataset, outperforming the second best by 4.7%.
Successful active speaker detection requires a three-stage pipeline: (i) audio-visual encoding for all speakers in the clip, (ii) inter-speaker relation modeling between a reference speaker and the background speakers within each frame, and (iii) temporal modeling for the reference speaker. Each stage of this pipeline plays an important role for the final performance of the created architecture. Based on a series of controlled experiments, this work presents several practical guidelines for audio-visual active speaker detection. Correspondingly, we present a new architecture called ASDNet, which achieves a new state-of-the-art on the AVA-ActiveSpeaker dataset with a mAP of 93.5% outperforming the second best with a large margin of 4.7%. Our code and pretrained models are publicly available.