Abstract:Balancing the trade-off between accuracy and speed for obtaining higher performance without sacrificing the inference time is a challenging topic for object detection task. Knowledge distillation, which serves as a kind of model compression techniques, provides a potential and feasible way to handle above efficiency and effectiveness issue through transferring the dark knowledge from the sophisticated teacher detector to the simple student one. Despite demonstrating promising solutions to make harmonies between accuracy and speed, current knowledge distillation for object detection methods still suffer from two limitations. Firstly, most of the methods are inherited or refereed from the frameworks in image classification task, and deploy an implicit manner by imitating or constraining the features from the intermediate layers or the output predictions between the teacher and student models. While little consideration has been raised to the intrinsic relevance of the classification and localization predictions in object detection task. Besides, these methods fail to investigate the relationship between detection and distillation tasks in knowledge distillation pipeline, and they train the whole network by simply integrating losses from these two different tasks through hand-crafted designation parameters. For addressing the aforementioned issues, we propose a novel Relation Knowledge Distillation by Auxiliary Learning for Object Detection (ReAL) method in this paper. Specifically, we first design a prediction relation distillation module which makes the student model directly mimic the output predictions from the teacher one, and conduct self and mutual relation distillation losses to excavate the relation information between teacher and student models. Moreover, for better devolving into the relationship between different tasks in distillation pipeline, we introduce the auxiliary learning into knowledge distillation for object detection and develop a dynamic weight adaptation strategy. Through regarding detection task as primary task and treating distillation task as auxiliary task in auxiliary learning framework, we dynamically adjust and regularize the corresponding weights of the losses for these tasks during the training process. Experiments on MS COCO dataset are conducted using various detector combinations of teacher and student models and the results show that our proposed ReAL can achieve obvious improvement on different distillation model configurations, while performing favorably against state-of-the-arts.

Improving Object Detection by Label Assignment Distillation

Research on Knowledge Distillation Algorithm of Object Detection

Learning Efficient Detector with Semi-supervised Adaptive Distillation

Improve Object Detection with Feature-based Knowledge Distillation: Towards Accurate and Efficient Detectors.

Prediction-Guided Distillation for Dense Object Detection

Learning Lightweight Object Detectors via Multi-Teacher Progressive Distillation

Task-Balanced Distillation for Object Detection

Knowledge Distillation for Object Detection via Rank Mimicking and Prediction-Guided Feature Imitation

Structured Knowledge Distillation for Accurate and Efficient Object Detection

Make a Strong Teacher with Label Assistance: A Novel Knowledge Distillation Approach for Semantic Segmentation

Focal and Global Knowledge Distillation for Detectors

Distilling Object Detectors via Decoupled Features

A dynamic label assignment strategy for one-stage detectors

LabelDistill: Label-guided Cross-modal Knowledge Distillation for Camera-based 3D Object Detection

Instance-Conditional Knowledge Distillation for Object Detection

Distilling Object Detectors With Fine-Grained Feature Imitation

Efficient Teacher: Semi-Supervised Object Detection for YOLOv5

Humble Teachers Teach Better Students for Semi-Supervised Object Detection

A New Multinetwork Mean Distillation Loss Function for Open-World Domain Incremental Object Detection

Label Matching Semi-Supervised Object Detection

Relation Knowledge Distillation by Auxiliary Learning for Object Detection