Rethinking and Improving the Robustness of Image Style Transfer

Abstract

Extensive research in neural style transfer methods has shown that the correlation between features extracted by a pre-trained VGG network has a remarkable ability to capture the visual style of an image. Surprisingly, however, this stylization quality is not robust and often degrades significantly when applied to features from more advanced and lightweight networks, such as those in the ResNet family. By performing extensive experiments with different network architectures, we find that residual connections, which represent the main architectural difference between VGG and ResNet, produce feature maps of small entropy, which are not suitable for style transfer. To improve the robustness of the ResNet architecture, we then propose a simple yet effective solution based on a softmax transformation of the feature activations that enhances their entropy. Experimental results demonstrate that this small magic can greatly improve the quality of stylization results, even for networks with random weights. This suggests that the architecture used for feature extraction is more important than the use of learned weights for the task of style transfer.

arXiv commentPublished in CVPR2021 (Oral)

References in corpus (10)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Distilling the Knowledge in a Neural Network
• Wide Residual Networks
• Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
• Sanity Checks for Saliency Maps
• Texture Synthesis Using Convolutional Neural Networks
• Learning Linear Transformations for Fast Arbitrary Style Transfer
• StyleBank: An Explicit Representation for Neural Image Style Transfer
• A Closed-form Solution to Photorealistic Image Stylization
• Attention-aware Multi-stroke Style Transfer