# Visual Person Understanding through Multi-Task and Multi-Dataset Learning

## TL;DR

This paper presents a multi-task learning approach for comprehensive person understanding, combining multiple datasets to improve performance across re-identification, attribute classification, body segmentation, and pose estimation.

## Contribution

It introduces a method to jointly learn multiple person-related tasks from diverse datasets without performance loss, suitable for resource-limited environments.

## Key findings

- Multi-task learning improves overall accuracy.
- Combining datasets enhances model robustness.
- Shared parameters do not significantly increase computational cost.

## Abstract

We address the problem of learning a single model for person re-identification, attribute classification, body part segmentation, and pose estimation. With predictions for these tasks we gain a more holistic understanding of persons, which is valuable for many applications. This is a classical multi-task learning problem. However, no dataset exists that these tasks could be jointly learned from. Hence several datasets need to be combined during training, which in other contexts has often led to reduced performance in the past. We extensively evaluate how the different task and datasets influence each other and how different degrees of parameter sharing between the tasks affect performance. Our final model matches or outperforms its single-task counterparts without creating significant computational overhead, rendering it highly interesting for resource-constrained scenarios such as mobile robotics.