SRNSD: Structure-Regularized Night-time Self-supervised Monocular Depth Estimation for Outdoor Scenes

人工智能 计算机视觉 单眼 计算机科学 模式识别(心理学)
作者
Runmin Cong,Chunlei Wu,Xibin Song,Wei Zhang,Sam Kwong,Hongdong Li,Pan Ji
出处
期刊:IEEE transactions on image processing [Institute of Electrical and Electronics Engineers]
卷期号:: 1-1
标识
DOI:10.1109/tip.2024.3465034
摘要

Deep CNNs have achieved impressive improvements for night-time self-supervised depth estimation form a monocular image. However, the performance degrades considerably compared to day-time depth estimation due to significant domain gaps, low visibility, and varying illuminations between day and night images. To address these challenges, we propose a novel night-time self-supervised monocular depth estimation framework with structure regularization, i.e., SRNSD, which incorporates three aspects of constraints for better performance, including feature and depth domain adaptation, image perspective constraint, and cropped multi-scale consistency loss. Specifically, we utilize adaptations of both feature and depth output spaces for better night-time feature extraction and depth map prediction, along with high- and low-frequency decoupling operations for better depth structure and texture recovery. Meanwhile, we employ an image perspective constraint to enhance the smoothness and obtain better depth maps in areas where the luminosity jumps change. Furthermore, we introduce a simple yet effective cropped multi-scale consistency loss that utilizes consistency among different scales of depth outputs for further optimization, refining the detailed textures and structures of predicted depth. Experimental results on different benchmarks with depth ranges of 40m and 60m, including Oxford RobotCar dataset, nuScenes dataset and CARLA-EPE dataset, demonstrate the superiority of our approach over state-of-the-art night-time self-supervised depth estimation approaches across multiple metrics, proving our effectiveness.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
qiao发布了新的文献求助10
1秒前
雷打不动我雷哥完成签到,获得积分10
1秒前
机灵的成协完成签到,获得积分10
1秒前
2秒前
Albert完成签到,获得积分10
2秒前
2秒前
2秒前
和谐的如柏完成签到,获得积分10
2秒前
4秒前
田様应助陈醒醒采纳,获得10
4秒前
5秒前
超级白昼完成签到,获得积分20
5秒前
云落发布了新的文献求助30
6秒前
瞬间发布了新的文献求助10
6秒前
7秒前
8秒前
8秒前
sorawing完成签到,获得积分10
8秒前
领导范儿应助dbndlk采纳,获得10
9秒前
9秒前
正经科研人完成签到,获得积分10
9秒前
月影发布了新的文献求助10
9秒前
汉堡包应助nmr采纳,获得10
10秒前
小马甲应助xiaowu采纳,获得10
11秒前
17发布了新的文献求助10
11秒前
yin完成签到,获得积分10
12秒前
华仔应助小月先先贝采纳,获得10
12秒前
13秒前
13秒前
zjiahao发布了新的文献求助10
14秒前
精灵夜雨发布了新的文献求助10
14秒前
oookkay完成签到,获得积分10
14秒前
jevon应助AmyDong采纳,获得10
15秒前
recognize完成签到,获得积分10
16秒前
17秒前
CipherSage应助17采纳,获得10
17秒前
搜集达人应助感性的俊驰采纳,获得10
19秒前
dbndlk完成签到,获得积分10
20秒前
20秒前
chenn发布了新的文献求助10
20秒前
高分求助中
歯科矯正学 第7版(或第5版) 1004
Semiconductor Process Reliability in Practice 1000
Smart but Scattered: The Revolutionary Executive Skills Approach to Helping Kids Reach Their Potential (第二版) 1000
GROUP-THEORY AND POLARIZATION ALGEBRA 500
Mesopotamian divination texts : conversing with the gods : sources from the first millennium BCE 500
Days of Transition. The Parsi Death Rituals(2011) 500
The Heath Anthology of American Literature: Early Nineteenth Century 1800 - 1865 Vol. B 500
热门求助领域 (近24小时)
化学 医学 生物 材料科学 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 基因 遗传学 催化作用 物理化学 免疫学 量子力学 细胞生物学
热门帖子
关注 科研通微信公众号,转发送积分 3233445
求助须知:如何正确求助?哪些是违规求助? 2879969
关于积分的说明 8213423
捐赠科研通 2547415
什么是DOI,文献DOI怎么找? 1376927
科研通“疑难数据库(出版商)”最低求助积分说明 647713
邀请新用户注册赠送积分活动 623150