DeepSocNav: Social Navigation by Imitating Human Behaviors

TL;DR

DeepSocNav introduces a novel approach to social navigation by generating synthetic first-person depth data from existing datasets using game engines, and employs a self-supervised task to improve navigation performance.

Contribution

The paper presents a method to convert bird's-eye view datasets into first-person depth views using Unity, enabling large-scale synthetic data generation for training social navigation models.

Findings

DeepSocNav outperforms baseline models in social navigation scores.

Synthetic data improves model training and generalization.

Self-supervised prediction of future depth frames enhances navigation accuracy.

Abstract

Current datasets to train social behaviors are usually borrowed from surveillance applications that capture visual data from a bird's-eye perspective. This leaves aside precious relationships and visual cues that could be captured through a first-person view of a scene. In this work, we propose a strategy to exploit the power of current game engines, such as Unity, to transform pre-existing bird's-eye view datasets into a first-person view, in particular, a depth view. Using this strategy, we are able to generate large volumes of synthetic data that can be used to pre-train a social navigation model. To test our ideas, we present DeepSocNav, a deep learning based model that takes advantage of the proposed approach to generate synthetic data. Furthermore, DeepSocNav includes a self-supervised strategy that is included as an auxiliary task. This consists of predicting the next depth frame that the agent will face. Our experiments show the benefits of the proposed model that is able to outperform relevant baselines in terms of social navigation scores.