Transferring Knowledge with Attention Distillation for Multi-Domain Image-to-Image Translation

Abstract

Gradient-based attention modeling has been used widely as a way to visualize and understand convolutional neural networks. However, exploiting these visual explanations during the training of generative adversarial networks (GANs) is an unexplored area in computer vision research. Indeed, we argue that this kind of information can be used to influence GANs training in a positive way. For this reason, in this paper, it is shown how gradient based attentions can be used as knowledge to be conveyed in a teacher-student paradigm for multi-domain image-to-image translation tasks in order to improve the results of the student architecture. Further, it is demonstrated how "pseudo"-attentions can also be employed during training when teacher and student networks are trained on different domains which share some similarities. The approach is validated on multi-domain facial attributes transfer and human expression synthesis showing both qualitative and quantitative results.

arXiv commentPreprint

References in corpus (9)

• Distilling the Knowledge in a Neural Network
• Conditional Generative Adversarial Nets
• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• Large Scale GAN Training for High Fidelity Natural Image Synthesis
• Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networks
• Self-Attention Generative Adversarial Networks
• Model compression via distillation and quantization
• Compressing GANs using Knowledge Distillation
• RelGAN: Multi-Domain Image-to-Image Translation via Relative Attributes