MMLatch: Bottom-up Top-down Fusion for Multimodal Sentiment Analysis
This work addresses a gap in deep learning models for multimodal sentiment analysis by incorporating top-down fusion inspired by human perception, offering incremental improvements for researchers and practitioners in affective computing.
The authors tackled the problem of missing top-down cross-modal interactions in multimodal sentiment analysis by proposing a neural architecture with a feedback mechanism that masks sensory inputs using high-level representations. Their method achieved state-of-the-art results on the CMU-MOSEI dataset, showing consistent improvements over established baselines like MulT.
Current deep learning approaches for multimodal fusion rely on bottom-up fusion of high and mid-level latent modality representations (late/mid fusion) or low level sensory inputs (early fusion). Models of human perception highlight the importance of top-down fusion, where high-level representations affect the way sensory inputs are perceived, i.e. cognition affects perception. These top-down interactions are not captured in current deep learning models. In this work we propose a neural architecture that captures top-down cross-modal interactions, using a feedback mechanism in the forward pass during network training. The proposed mechanism extracts high-level representations for each modality and uses these representations to mask the sensory inputs, allowing the model to perform top-down feature masking. We apply the proposed model for multimodal sentiment recognition on CMU-MOSEI. Our method shows consistent improvements over the well established MulT and over our strong late fusion baseline, achieving state-of-the-art results.