Abstract:Minimizing the computation complexity is essential for the popularization of deep networks in practical applications. Nowadays, most researches attempt to accelerate deep networks by designing new network structure or compressing the network parameters. Meanwhile, transfer learning techniques such as knowledge distillation are utilized to keep the performance of deep models. In this paper, we focus on accelerating deep models and relieving the computation burden by using low-resolution (LR) images as inputs while maintaining competitive performance, which is rarely researched in the current literature. Deep networks may encounter serious performance degradation when using LR inputs because many details are unavailable from LR images. Besides, the existing approaches may fail to learn discriminative features for LR images because of the dramatic appearance variations between LR and high-resolution (HR) images. To tackle with the above problems, we propose a resolution-aware knowledge distillation (RKD) framework to narrow the cross-resolution variations by transferring knowledge from HR domain to LR domain. The proposed framework consists of a HR teacher network and a LR student network. First, we introduce a discriminator and propose an adversarial learning strategy to shrink the variations between inputs with changing resolution. Then we design a cross-resolution knowledge distillation (CRKD) loss to train discriminative student network by exploiting the knowledge of the teacher network. The CRKD loss is consisted of a resolution-aware distillation loss, a pair-wise constraint, and a maximum mean discrepancy loss. Experimental results on person re-identification, image classification, face recognition, and defect segmentation tasks demonstrate that RKD outperforms traditional knowledge distillation method by achieving better performance with lower computation complexities. Furthermore, CRKD surpasses the state-of-the-art knowledge distillation methods in transferring knowledge across different resolutions under RKD framework, especially when coping with large resolution differences.

Look One and More: Distilling Hybrid Order Relational Knowledge for Cross-Resolution Image Recognition

Low-Resolution Object Recognition with Cross-Resolution Relational Contrastive Distillation

Low-Resolution Visual Recognition Via Deep Feature Distillation

Low-Resolution Face Recognition via Adaptable Instance-Relation Distillation

Multi-level knowledge distillation for low-resolution object detection and facial expression recognition

Cross-resolution Face Recognition via Identity-Preserving Network and Knowledge Distillation

Low-Resolution Face Recognition in the Wild Via Selective Knowledge Distillation

Resolution-Aware Knowledge Distillation for Efficient Inference.

Efficient Low-Resolution Face Recognition via Bridge Distillation

Pixel Distillation: A New Knowledge Distillation Scheme for Low-Resolution Image Recognition

One-stage Low-resolution Text Recognition with High-resolution Knowledge Transfer

Cross-Domain Knowledge Distillation for Low-Resolution Human Pose Estimation

Hybrid mix-up contrastive knowledge distillation

Distilling Generative-Discriminative Representations for Very Low-Resolution Face Recognition

Multistage feature fusion knowledge distillation

Information Fusion with Knowledge Distillation for Fine-grained Remote Sensing Object Detection

Distilling Knowledge from Large-Scale Image Models for Object Detection

Expanding and Refining Hybrid Compressors for Efficient Object Re-Identification

Super-resolution Guided Knowledge Distillation for Low-Resolution Image Classification

Learning Lightweight Object Detectors via Multi-Teacher Progressive Distillation

A Multi-Teacher Assisted Knowledge Distillation Approach for Enhanced Face Image Authentication.