Multi-scale Attention Guided Pose Transfer

TL;DR

This paper introduces an enhanced multi-scale attention network for pose transfer, significantly improving visual quality and analytical performance over existing methods by integrating dense attention links at all resolution levels.

Contribution

It proposes a novel dense multi-scale attention guided architecture with attention links at every resolution level for improved pose transfer results.

Findings

Significant visual quality improvements over prior methods.

Quantitative metrics show enhanced performance.

Extensive comparisons validate the approach.

Abstract

Pose transfer refers to the probabilistic image generation of a person with a previously unseen novel pose from another image of that person having a different pose. Due to potential academic and commercial applications, this problem is extensively studied in recent years. Among the various approaches to the problem, attention guided progressive generation is shown to produce state-of-the-art results in most cases. In this paper, we present an improved network architecture for pose transfer by introducing attention links at every resolution level of the encoder and decoder. By utilizing such dense multi-scale attention guided approach, we are able to achieve significant improvement over the existing methods both visually and analytically. We conclude our findings with extensive qualitative and quantitative comparisons against several existing methods on the DeepFashion dataset.