Saved in:
Bibliographic Details
Main Authors: Chen, Xi, Cai, Yang, Wu, Yuan, Xiong, Bo, Park, Taesung
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.04618
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911772722593792
author Chen, Xi
Cai, Yang
Wu, Yuan
Xiong, Bo
Park, Taesung
author_facet Chen, Xi
Cai, Yang
Wu, Yuan
Xiong, Bo
Park, Taesung
contents Recently, MBConv blocks, initially designed for efficiency in resource-limited settings and later adapted for cutting-edge image classification performances, have demonstrated significant potential in image classification tasks. Despite their success, their application in semantic segmentation has remained relatively unexplored. This paper introduces a novel adaptation of MBConv blocks specifically tailored for semantic segmentation. Our modification stems from the insight that semantic segmentation requires the extraction of more detailed spatial information than image classification. We argue that to effectively perform multi-scale semantic segmentation, each branch of a U-Net architecture, regardless of its resolution, should possess equivalent segmentation capabilities. By implementing these changes, our approach achieves impressive mean Intersection over Union (IoU) scores of 84.5% and 84.0% on the Cityscapes test and validation datasets, respectively, demonstrating the efficacy of our proposed modifications in enhancing semantic segmentation performance.
format Preprint
id arxiv_https___arxiv_org_abs_2402_04618
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Multi-Scale Semantic Segmentation with Modified MBConv Blocks
Chen, Xi
Cai, Yang
Wu, Yuan
Xiong, Bo
Park, Taesung
Computer Vision and Pattern Recognition
Recently, MBConv blocks, initially designed for efficiency in resource-limited settings and later adapted for cutting-edge image classification performances, have demonstrated significant potential in image classification tasks. Despite their success, their application in semantic segmentation has remained relatively unexplored. This paper introduces a novel adaptation of MBConv blocks specifically tailored for semantic segmentation. Our modification stems from the insight that semantic segmentation requires the extraction of more detailed spatial information than image classification. We argue that to effectively perform multi-scale semantic segmentation, each branch of a U-Net architecture, regardless of its resolution, should possess equivalent segmentation capabilities. By implementing these changes, our approach achieves impressive mean Intersection over Union (IoU) scores of 84.5% and 84.0% on the Cityscapes test and validation datasets, respectively, demonstrating the efficacy of our proposed modifications in enhancing semantic segmentation performance.
title Multi-Scale Semantic Segmentation with Modified MBConv Blocks
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2402.04618