Dimensions overall score 7.0

• The proposed method simultaneously achieves group detection and shared attention estimation using a two-step process.method 95%
• Most previous methods estimate shared attention without detecting the actual group of people focusing on it.limitation 90%
• Most previous methods assume there is a single shared attention point in a given image.limitation 90%
• The method outperforms other methods in group detection and shared attention estimation.result 85%
• The method uses a refinement step that accounts for initial SA heatmaps to improve group membership predictions.method 85%
• The method generates SA heatmaps relying on individual gaze attention heatmaps and group membership scalars.method 90%
• The limitations of previous methods impact performance and limit practical applicability.limitation 80%
• Additional analyses validate the effectiveness of the proposed components.result 80%

Competitive landscape

Competitor map is still being generated for this paper. Enable generation or check back soon.

Startup potential card