Exemplar-based Generative Facial Editing

TL;DR

This paper introduces a novel exemplar-based facial editing method using region inpainting, which enables diverse, personalized, and controllable face edits by learning from reference images while avoiding undesired attribute transfer.

Contribution

The proposed approach uniquely combines region inpainting with attribute label constraints to improve control and personalization in facial editing tasks.

Findings

Produces diverse and personalized face edits

Offers greater user control than existing methods

Effectively prevents undesired attribute transfer

Abstract

Image synthesis has witnessed substantial progress due to the increasing power of generative model. This paper we propose a novel generative approach for exemplar based facial editing in the form of the region inpainting. Our method first masks the facial editing region to eliminates the pixel constraints of the original image, then exemplar based facial editing can be achieved by learning the corresponding information from the reference image to complete the masked region. In additional, we impose the attribute labels constraint to model disentangled encodings in order to avoid undesired information being transferred from the exemplar to the original image editing region. Experimental results demonstrate our method can produce diverse and personalized face editing results and provide far more user control flexibility than nearly all existing methods.