Abstract:Existing image-to-image translation methods usually adopt an encoder-decoder structure to generate images. The encoder extracts the features of input images using a sequence of convolution layers until a bottleneck, and then, the intermediate features are decoded to the target image. However, the existence of bottleneck layer in such structure may lead to blurry and bad quality of the translated images, since different domain translations may be related to the global or local region in the input image or even in an abstract level. To prevent these problems, we propose the channel attention networks for image translation in this paper. It is a novel model that supports the multi-domain image-to-image translation using one single model. Conditioning on the target domain label, an auto-encoder-like network with multiple attention connections is trained to translate the input image into the target domain The attention connections better shuttle the low-level information in the encoder to the decoder, which helps to preserve the structure. A multi-level attention mechanism is also designed in the proposed model to further improve the performance of our model. More specially, the feature maps in the encoder are first squeezed by average pooling and used to output a channel-wise attention mask. The attention mask softly determines which channels of the feature maps are translated and which channels are kept. By enforcing the model to learn a cyclic domain transformation during training, our model does not require paired training data, which greatly improves the versatility to different kinds of data. We experimentally demonstrated the effectiveness of our proposed model on the facial and clothing image translation tasks. The extensive ablations are also conducted to further validate the contribution of the proposed attention module used in our model.

Contrastive Learning with Attention Mechanism and Multi-Scale Sample Network for Unpaired Image-to-Image Translation

Unpaired Salient Object Translation Via Spatial Attention Prior

Multimodal Image-to-Image Translation via Mutual Information Estimation and Maximization

Image Cross-Domain Translation Algorithm Based on Self-Similarity and Contrastive Learning

AttentionGAN: Unpaired Image-to-Image Translation Using Attention-Guided Generative Adversarial Networks

Channel Attention Networks for Image Translation.

Multilevel Contrast Strategy for Unpaired Image-to-image Translation

Multi-attention bidirectional contrastive learning method for unpaired image-to-image translation

Unsupervised image-to-image translation with multiscale attention generative adversarial network

Show, Attend and Translate: Unpaired Multi-Domain Image-to-Image Translation with Visual Attention

A one-to-many conditional generative adversarial network framework for multiple image-to-image translations

Global and Local Alignment Networks for Unpaired Image-to-Image Translation

Unsupervised Object-Level Image-to-Image Translation Using Positional Attention Bi-Flow Generative Network.

CSAGAN: Channel and Spatial Attention-Guided Generative Adversarial Networks for Unsupervised Image-to-Image Translation

Image Translation with Attention Mechanism Based on Generative Adversarial Networks.

Multi-cropping Contrastive Learning and Domain Consistency for Unsupervised Image-to-Image Translation

Single-image Translation Based on Improved Spatial Attention and Hybrid Convolution

Edge-guided Adversarial Network Based on Contrastive Learning for Image-to-Image Translation

Image-to-image Translation Model Combining GAN and Multi-angle Attention

HE-ASR-IT: Hybrid Excitation and Adaptive Style Recombination for Unpaired Image-to-Image Translation

A Unified Hyper-GAN Model for Unpaired Multi-contrast MR Image Translation