Instance-Conditional Knowledge Distillation for Object Detection

Knowledge distillation has shown great success in classification, however, it is still challenging for detection. In a typical image for detection, representations from different locations may have different contributions to detection targets, making the distillation hard to balance. In this paper, we propose a conditional distillation framework to distill the desired knowledge, namely knowledge that is beneficial in terms of both classification and localization for every instance. The framework introduces a learnable conditional decoding module, which retrieves information given each target instance as query. Specifically, we encode the condition information as query and use the teacher's representations as key. The attention between query and key is used to measure the contribution of different features, guided by a localization-recognition-sensitive auxiliary task. Extensive experiments demonstrate the efficacy of our method: we observe impressive improvements under various settings. Notably, we boost RetinaNet with ResNet-50 backbone from $37. 4$ to $40. 7$ mAP ($+3. 3$) under $1\times$ schedule, that even surpasses the teacher ($40. 4$ mAP) with ResNet-101 backbone under $3\times$ schedule. Code has been released on https: //github. com/megvii-research/ICD.

Authors

• Zijian Kang
• Peizhen Zhang
• Xiangyu Zhang Megvii Technology Beijing Academy of Artifcial Intelligence
• Jian Sun Xi'an Jiaotong University Pazhou Laboratory (Huangpu), China Peng Cheng Laboratory, China
• Nanning Zheng National Key Laboratory of Human-Machine Hybrid Augmented Intelligence National Engineering Research Center of Visual Information and Applications Institute of Artificial Intelligence and Robotics, Xi’an Jiaotong University

Keywords

No keywords are indexed for this paper.