Investigating Modality Bias in Audio Visual Video Parsing
This addresses a specific problem in weakly supervised video analysis for researchers, but it is incremental as it builds on an existing architecture.
The paper tackled modality bias in audio-visual video parsing, where existing models ignore certain modalities, and proposed a feature aggregation variant that achieved absolute F-score gains of about 2% for visual events and 1.6% for audio-visual events.
We focus on the audio-visual video parsing (AVVP) problem that involves detecting audio and visual event labels with temporal boundaries. The task is especially challenging since it is weakly supervised with only event labels available as a bag of labels for each video. An existing state-of-the-art model for AVVP uses a hybrid attention network (HAN) to generate cross-modal features for both audio and visual modalities, and an attentive pooling module that aggregates predicted audio and visual segment-level event probabilities to yield video-level event probabilities. We provide a detailed analysis of modality bias in the existing HAN architecture, where a modality is completely ignored during prediction. We also propose a variant of feature aggregation in HAN that leads to an absolute gain in F-scores of about 2% and 1.6% for visual and audio-visual events at both segment-level and event-level, in comparison to the existing HAN model.