no code implementations • 21 Aug 2022 • Junghun Kim, Jihie Kim
In particular, we propose a cosine similarity-based graph as an ideal graph structure for representation learning in SER.
no code implementations • 21 Aug 2022 • Junghun Kim, Jihie Kim
In this paper, we present a Context-driven Modality Shifting BERT with Contrastive Learning for linguistic, visual, acoustic Representations (CMSBERT-CLR), which incorporates the whole context's non-verbal and verbal information and aligns modalities more effectively through contrastive learning.
no code implementations • 21 Aug 2022 • Junghun Kim, Yoojin An, Jihie Kim
To improve the attention area, we propose to use a Focus-Attention (FA) mechanism and a novel Calibration-Attention (CA) mechanism in combination with the multi-head self-attention.