# Towards Social Artificial Intelligence: Nonverbal Social Signal Prediction in A Triadic Interaction

## TL;DR

This paper introduces a new dataset and task for predicting social signals like body, face, and hand motions in triadic interactions, aiming to enable machines to better understand and communicate socially with humans.

## Contribution

It formulates the social signal prediction problem and provides a novel 3D motion capture dataset for modeling social interactions in a data-driven manner.

## Key findings

- Baseline models can predict speaking status, social formation, and gestures.
- The dataset captures a broad spectrum of social signals in triadic interactions.
- This work advances the development of socially aware artificial intelligence.

## Abstract

We present a new research task and a dataset to understand human social interactions via computational methods, to ultimately endow machines with the ability to encode and decode a broad channel of social signals humans use. This research direction is essential to make a machine that genuinely communicates with humans, which we call Social Artificial Intelligence. We first formulate the "social signal prediction" problem as a way to model the dynamics of social signals exchanged among interacting individuals in a data-driven way. We then present a new 3D motion capture dataset to explore this problem, where the broad spectrum of social signals (3D body, face, and hand motions) are captured in a triadic social interaction scenario. Baseline approaches to predict speaking status, social formation, and body gestures of interacting individuals are presented in the defined social prediction framework.