# Graphonomy: Universal Human Parsing via Graph Transfer Learning

## TL;DR

Graphonomy introduces a universal human parsing model that leverages hierarchical graph transfer learning to unify diverse label granularities and domains, achieving state-of-the-art results across multiple benchmarks.

## Contribution

The paper proposes Graphonomy, a novel hierarchical graph transfer learning framework that encodes semantic label structures and transfers knowledge across datasets for universal human parsing.

## Key findings

- Achieves state-of-the-art results on three benchmarks.
- Effectively unifies human parsing across diverse datasets.
- Demonstrates superior generalization to various parsing tasks.

## Abstract

Prior highly-tuned human parsing models tend to fit towards each dataset in a specific domain or with discrepant label granularity, and can hardly be adapted to other human parsing tasks without extensive re-training. In this paper, we aim to learn a single universal human parsing model that can tackle all kinds of human parsing needs by unifying label annotations from different domains or at various levels of granularity. This poses many fundamental learning challenges, e.g. discovering underlying semantic structures among different label granularity, performing proper transfer learning across different image domains, and identifying and utilizing label redundancies across related tasks. To address these challenges, we propose a new universal human parsing agent, named "Graphonomy", which incorporates hierarchical graph transfer learning upon the conventional parsing network to encode the underlying label semantic structures and propagate relevant semantic information. In particular, Graphonomy first learns and propagates compact high-level graph representation among the labels within one dataset via Intra-Graph Reasoning, and then transfers semantic information across multiple datasets via Inter-Graph Transfer. Various graph transfer dependencies (\eg, similarity, linguistic knowledge) between different datasets are analyzed and encoded to enhance graph transfer capability. By distilling universal semantic graph representation to each specific task, Graphonomy is able to predict all levels of parsing labels in one system without piling up the complexity. Experimental results show Graphonomy effectively achieves the state-of-the-art results on three human parsing benchmarks as well as advantageous universal human parsing performance.