Self-Learning with Rectification Strategy for Human Parsing

TL;DR

This paper introduces a self-learning approach with a trainable graph reasoning strategy to improve human parsing accuracy by correcting pseudo-label errors, leveraging global structural and local consistency modules.

Contribution

It proposes a novel trainable graph reasoning method that rectifies pseudo-label errors by modeling human body topology and pixel relations, enhancing self-learning in human parsing.

Findings

Outperforms state-of-the-art methods on LIP and ATR datasets.

Effective correction of global and local pseudo-label errors.

Improved human parsing accuracy with self-learning strategy.

Abstract

In this paper, we solve the sample shortage problem in the human parsing task. We begin with the self-learning strategy, which generates pseudo-labels for unlabeled data to retrain the model. However, directly using noisy pseudo-labels will cause error amplification and accumulation. Considering the topology structure of human body, we propose a trainable graph reasoning method that establishes internal structural connections between graph nodes to correct two typical errors in the pseudo-labels, i.e., the global structural error and the local consistency error. For the global error, we first transform category-wise features into a high-level graph model with coarse-grained structural information, and then decouple the high-level graph to reconstruct the category features. The reconstructed features have a stronger ability to represent the topology structure of the human body. Enlarging the receptive field of features can effectively reducing the local error. We first project feature pixels into a local graph model to capture pixel-wise relations in a hierarchical graph manner, then reverse the relation information back to the pixels. With the global structural and local consistency modules, these errors are rectified and confident pseudo-labels are generated for retraining. Extensive experiments on the LIP and the ATR datasets demonstrate the effectiveness of our global and local rectification modules. Our method outperforms other state-of-the-art methods in supervised human parsing tasks.