Region-aware Knowledge Distillation for Efficient Image-to-Image Translation
원문 초록
Recent progress in image-to-image translation has witnessed the success of generative adversarial networks (GANs). However, GANs usually contain a huge number of parameters, which lead to intolerant memory and computation consumption and limit their deployment on edge devices. To address this issue, knowledge distillation is proposed to transfer the knowledge from a cumbersome teacher model to an efficient student model. However, most previous knowledge distillation methods are designed for image classification and lead to limited performance in image-to-image translation. In this paper, we propose Region-aware Knowledge Distillation ReKo to compress image-to-image translation models. Firstly, ReKo adaptively finds the crucial regions in the images with an attention module. Then, patch-wise contrastive learning is adopted to maximize the mutual information between students and teachers in these crucial regions. Experiments with eight comparison methods on nine datasets demonstrate the substantial effectiveness of ReKo on both paired and unpaired image-to-image translation. For instance, our 7.08X compressed and 6.80X accelerated CycleGAN student outperforms its teacher by 1.33 and 1.04 FID scores on Horse to Zebra and Zebra to Horse, respectively. Codes will be released on GitHub.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
SeqPE: Transformer with Sequential Position Encoding
PiPViT: Patch-based Visual Interpretable Prototypes for Retinal Image Analysis
Iterative Quantum Feature Maps
IPCL: Iterative Pseudo-Supervised Contrastive Learning to Improve Self-Supervised Feature Representation
A Unified Framework against Topology and Class Imbalance
Head Network Distillation: Splitting Distilled Deep Neural Networks for Resource-Constrained Edge Computing Systems
토론
이 논문과 연결된 포럼 글이 아직 없습니다.