Improving Social Awareness Through DANTE: A Deep Affinity Network for Clustering Conversational Interactants
This work addresses the problem of social awareness for applications like robotics, though it appears incremental by combining deep learning with classical clustering.
The paper tackles the problem of detecting conversational groups by identifying spatial arrangements, using a Deep Affinity Network (DANTE) to predict pair-wise affinities and graph clustering to identify groups. The results show improved detection on established benchmarks compared to prior approaches, with practical application in human-robot interaction.
We propose a data-driven approach to detect conversational groups by identifying spatial arrangements typical of these focused social encounters. Our approach uses a novel Deep Affinity Network (DANTE) to predict the likelihood that two individuals in a scene are part of the same conversational group, considering their social context. The predicted pair-wise affinities are then used in a graph clustering framework to identify both small (e.g., dyads) and large groups. The results from our evaluation on multiple, established benchmarks suggest that combining powerful deep learning methods with classical clustering techniques can improve the detection of conversational groups in comparison to prior approaches. Finally, we demonstrate the practicality of our approach in a human-robot interaction scenario. Our efforts show that our work advances group detection not only in theory, but also in practice.