Saved in:
Bibliographic Details
Main Authors: Pan, Yuwen, Sun, Rui, Luo, Naisong, Zhang, Tianzhu, Zhang, Yongdong
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.13838
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910576177840128
author Pan, Yuwen
Sun, Rui
Luo, Naisong
Zhang, Tianzhu
Zhang, Yongdong
author_facet Pan, Yuwen
Sun, Rui
Luo, Naisong
Zhang, Tianzhu
Zhang, Yongdong
contents Semantic segmentation of night-time images holds significant importance in computer vision, particularly for applications like night environment perception in autonomous driving systems. However, existing methods tend to parse night-time images from a day-time perspective, leaving the inherent challenges in low-light conditions (such as compromised texture and deceiving matching errors) unexplored. To address these issues, we propose a novel end-to-end optimized approach, named NightFormer, tailored for night-time semantic segmentation, avoiding the conventional practice of forcibly fitting night-time images into day-time distributions. Specifically, we design a pixel-level texture enhancement module to acquire texture-aware features hierarchically with phase enhancement and amplified attention, and an object-level reliable matching module to realize accurate association matching via reliable attention in low-light environments. Extensive experimental results on various challenging benchmarks including NightCity, BDD and Cityscapes demonstrate that our proposed method performs favorably against state-of-the-art night-time semantic segmentation methods.
format Preprint
id arxiv_https___arxiv_org_abs_2408_13838
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Exploring Reliable Matching with Phase Enhancement for Night-time Semantic Segmentation
Pan, Yuwen
Sun, Rui
Luo, Naisong
Zhang, Tianzhu
Zhang, Yongdong
Computer Vision and Pattern Recognition
Semantic segmentation of night-time images holds significant importance in computer vision, particularly for applications like night environment perception in autonomous driving systems. However, existing methods tend to parse night-time images from a day-time perspective, leaving the inherent challenges in low-light conditions (such as compromised texture and deceiving matching errors) unexplored. To address these issues, we propose a novel end-to-end optimized approach, named NightFormer, tailored for night-time semantic segmentation, avoiding the conventional practice of forcibly fitting night-time images into day-time distributions. Specifically, we design a pixel-level texture enhancement module to acquire texture-aware features hierarchically with phase enhancement and amplified attention, and an object-level reliable matching module to realize accurate association matching via reliable attention in low-light environments. Extensive experimental results on various challenging benchmarks including NightCity, BDD and Cityscapes demonstrate that our proposed method performs favorably against state-of-the-art night-time semantic segmentation methods.
title Exploring Reliable Matching with Phase Enhancement for Night-time Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.13838