Supervised Attribute Information Removal and Reconstruction for Image Manipulation
This addresses the issue of correlated attributes in image manipulation for applications like photo editing, though it is incremental as it builds on prior disentanglement methods.
The paper tackles the problem of unwanted image editing effects in attribute manipulation by proposing an Attribute Information Removal and Reconstruction (AIRR) network that removes attribute information entirely and injects desired attributes, improving attribute manipulation accuracy and top-k retrieval rate by 10% on average over prior work.
The goal of attribute manipulation is to control specified attribute(s) in given images. Prior work approaches this problem by learning disentangled representations for each attribute that enables it to manipulate the encoded source attributes to the target attributes. However, encoded attributes are often correlated with relevant image content. Thus, the source attribute information can often be hidden in the disentangled features, leading to unwanted image editing effects. In this paper, we propose an Attribute Information Removal and Reconstruction (AIRR) network that prevents such information hiding by learning how to remove the attribute information entirely, creating attribute excluded features, and then learns to directly inject the desired attributes in a reconstructed image. We evaluate our approach on four diverse datasets with a variety of attributes including DeepFashion Synthesis, DeepFashion Fine-grained Attribute, CelebA and CelebA-HQ, where our model improves attribute manipulation accuracy and top-k retrieval rate by 10% on average over prior work. A user study also reports that AIRR manipulated images are preferred over prior work in up to 76% of cases.