Social Attention

Social attention research focuses on how systems, including humans and robots, selectively attend to relevant information from multiple sources (e.g., visual, auditory, textual) in complex social contexts. Current research emphasizes improving attention mechanisms in models like Transformers, exploring novel algorithms such as expressive attention and human-like attention that prioritize relevant information and mitigate the impact of irrelevant or conflicting cues. This work is significant for advancing artificial intelligence, particularly in areas like human-robot interaction and natural language processing, by enabling more robust and contextually aware systems capable of handling complex multimodal inputs.