Distilling Inter-Class Distance for Semantic Segmentation
This work addresses the problem of efficient semantic segmentation for computer vision applications, offering an incremental improvement over existing distillation methods.
The paper tackles the problem of knowledge distillation for semantic segmentation by addressing the neglected transfer of inter-class distance knowledge, proposing an Inter-class Distance Distillation (IDD) method and a position information distillation module. The result shows improved accuracy, achieving state-of-the-art performance with a 7.50% boost on the Cityscapes dataset.
Knowledge distillation is widely adopted in semantic segmentation to reduce the computation cost.The previous knowledge distillation methods for semantic segmentation focus on pixel-wise feature alignment and intra-class feature variation distillation, neglecting to transfer the knowledge of the inter-class distance in the feature space, which is important for semantic segmentation. To address this issue, we propose an Inter-class Distance Distillation (IDD) method to transfer the inter-class distance in the feature space from the teacher network to the student network. Furthermore, semantic segmentation is a position-dependent task,thus we exploit a position information distillation module to help the student network encode more position information. Extensive experiments on three popular datasets: Cityscapes, Pascal VOC and ADE20K show that our method is helpful to improve the accuracy of semantic segmentation models and achieves the state-of-the-art performance. E.g. it boosts the benchmark model("PSPNet+ResNet18") by 7.50% in accuracy on the Cityscapes dataset.