Attention-guided Feature Distillation for Semantic Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mansourian, Amir M., Jalali, Arya, Ahmadi, Rozhan, Kasaei, Shohreh
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915209996664832
author Mansourian, Amir M.
Jalali, Arya
Ahmadi, Rozhan
Kasaei, Shohreh
author_facet Mansourian, Amir M.
Jalali, Arya
Ahmadi, Rozhan
Kasaei, Shohreh
contents Deep learning models have achieved significant results across various computer vision tasks. However, due to the large number of parameters in these models, deploying them in real-time scenarios is a critical challenge, specifically in dense prediction tasks such as semantic segmentation. Knowledge distillation has emerged as a successful technique for addressing this problem by transferring knowledge from a cumbersome model (teacher) to a lighter model (student). In contrast to existing complex methodologies commonly employed for distilling knowledge from a teacher to a student, this paper showcases the efficacy of a simple yet powerful method for utilizing refined feature maps to transfer attention. The proposed method has proven to be effective in distilling rich information, outperforming existing methods in semantic segmentation as a dense prediction task. The proposed Attention-guided Feature Distillation (AttnFD) method, employs the Convolutional Block Attention Module (CBAM), which refines feature maps by taking into account both channel-specific and spatial information content. Simply using the Mean Squared Error (MSE) loss function between the refined feature maps of the teacher and the student, AttnFD demonstrates outstanding performance in semantic segmentation, achieving state-of-the-art results in terms of improving the mean Intersection over Union (mIoU) of the student network on the PascalVoc 2012, Cityscapes, COCO, and CamVid datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2403_05451
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Attention-guided Feature Distillation for Semantic Segmentation
Mansourian, Amir M.
Jalali, Arya
Ahmadi, Rozhan
Kasaei, Shohreh
Computer Vision and Pattern Recognition
Deep learning models have achieved significant results across various computer vision tasks. However, due to the large number of parameters in these models, deploying them in real-time scenarios is a critical challenge, specifically in dense prediction tasks such as semantic segmentation. Knowledge distillation has emerged as a successful technique for addressing this problem by transferring knowledge from a cumbersome model (teacher) to a lighter model (student). In contrast to existing complex methodologies commonly employed for distilling knowledge from a teacher to a student, this paper showcases the efficacy of a simple yet powerful method for utilizing refined feature maps to transfer attention. The proposed method has proven to be effective in distilling rich information, outperforming existing methods in semantic segmentation as a dense prediction task. The proposed Attention-guided Feature Distillation (AttnFD) method, employs the Convolutional Block Attention Module (CBAM), which refines feature maps by taking into account both channel-specific and spatial information content. Simply using the Mean Squared Error (MSE) loss function between the refined feature maps of the teacher and the student, AttnFD demonstrates outstanding performance in semantic segmentation, achieving state-of-the-art results in terms of improving the mean Intersection over Union (mIoU) of the student network on the PascalVoc 2012, Cityscapes, COCO, and CamVid datasets.
title Attention-guided Feature Distillation for Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.05451