Conversation Group Detection With Spatio-Temporal Context

TL;DR

This paper introduces a deep learning approach using dynamic LSTM models and graph clustering to detect conversation groups in social scenarios from overhead video recordings, emphasizing temporal and spatial context.

Contribution

It presents a novel method leveraging temporal dynamics and continuous affinity predictions for improved conversation group detection in social settings.

Findings

Improved group detection accuracy with temporal granularity.

Affinity values correlate with conversation group membership.

Predicted affinities can forecast future group memberships.

Abstract

In this work, we propose an approach for detecting conversation groups in social scenarios like cocktail parties and networking events, from overhead camera recordings. We posit the detection of conversation groups as a learning problem that could benefit from leveraging the spatial context of the surroundings, and the inherent temporal context in interpersonal dynamics which is reflected in the temporal dynamics in human behavior signals, an aspect that has not been addressed in recent prior works. This motivates our approach which consists of a dynamic LSTM-based deep learning model that predicts continuous pairwise affinity values indicating how likely two people are in the same conversation group. These affinity values are also continuous in time, since relationships and group membership do not occur instantaneously, even though the ground truths of group membership are binary. Using the predicted affinity values, we apply a graph clustering method based on Dominant Set extraction to identify the conversation groups. We benchmark the proposed method against established methods on multiple social interaction datasets. Our results showed that the proposed method improves group detection performance in data that has more temporal granularity in conversation group labels. Additionally, we provide an analysis in the predicted affinity values in relation to the conversation group detection. Finally, we demonstrate the usability of the predicted affinity values in a forecasting framework to predict group membership for a given forecast horizon.