Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection

With the ease of acquiring RGB-D images from line-scan 3D cameras and the development of computer vision, anomaly detection is now widely applied to railway inspection. As 2D anomaly detection is susceptible to capturing condition, a combination of depth maps is now being explored in industrial insp...

Full description

Saved in:
Bibliographic Details
Main Authors: Xiaorong Gao, Pengxu Wen, Jinlong Li, Lin Luo
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/8/2631
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850180200263843840
author Xiaorong Gao
Pengxu Wen
Jinlong Li
Lin Luo
author_facet Xiaorong Gao
Pengxu Wen
Jinlong Li
Lin Luo
author_sort Xiaorong Gao
collection DOAJ
description With the ease of acquiring RGB-D images from line-scan 3D cameras and the development of computer vision, anomaly detection is now widely applied to railway inspection. As 2D anomaly detection is susceptible to capturing condition, a combination of depth maps is now being explored in industrial inspection to reduce these interferences. In this case, this paper proposes a novel approach for RGB-D anomaly detection called Dual-Branch Cross-Fusion Normalizing Flow (DCNF). In this work, we aim to exploit the fusion strategy for dual-branch normalizing flow with multi-modal inputs to be applied in the field of track detection. On the one hand, we introduce the mutual perception module to acquire cross-complementary prior knowledge in the early stage. On the other hand, we exploit the effectiveness of the fusion flow to fuse the dual-branch of RGB-D inputs. We experiment on the real-world Track Anomaly (TA) dataset. The performance evaluation of DCNF on TA dataset achieves an impressive AUROC score of 98.49%, which is 3.74% higher than the second-best method.
format Article
id doaj-art-b17eda17c96e45ddb5fee6e5ce7e16d4
institution OA Journals
issn 1424-8220
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-b17eda17c96e45ddb5fee6e5ce7e16d42025-08-20T02:18:15ZengMDPI AGSensors1424-82202025-04-01258263110.3390/s25082631Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly DetectionXiaorong Gao0Pengxu Wen1Jinlong Li2Lin Luo3School of Physical Science and Technology, Southwest Jiaotong University, Chengdu 610031, ChinaSchool of Physical Science and Technology, Southwest Jiaotong University, Chengdu 610031, ChinaSchool of Physical Science and Technology, Southwest Jiaotong University, Chengdu 610031, ChinaSchool of Physical Science and Technology, Southwest Jiaotong University, Chengdu 610031, ChinaWith the ease of acquiring RGB-D images from line-scan 3D cameras and the development of computer vision, anomaly detection is now widely applied to railway inspection. As 2D anomaly detection is susceptible to capturing condition, a combination of depth maps is now being explored in industrial inspection to reduce these interferences. In this case, this paper proposes a novel approach for RGB-D anomaly detection called Dual-Branch Cross-Fusion Normalizing Flow (DCNF). In this work, we aim to exploit the fusion strategy for dual-branch normalizing flow with multi-modal inputs to be applied in the field of track detection. On the one hand, we introduce the mutual perception module to acquire cross-complementary prior knowledge in the early stage. On the other hand, we exploit the effectiveness of the fusion flow to fuse the dual-branch of RGB-D inputs. We experiment on the real-world Track Anomaly (TA) dataset. The performance evaluation of DCNF on TA dataset achieves an impressive AUROC score of 98.49%, which is 3.74% higher than the second-best method.https://www.mdpi.com/1424-8220/25/8/2631anomaly detectionnormalizing flowRGB-D fusion
spellingShingle Xiaorong Gao
Pengxu Wen
Jinlong Li
Lin Luo
Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection
Sensors
anomaly detection
normalizing flow
RGB-D fusion
title Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection
title_full Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection
title_fullStr Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection
title_full_unstemmed Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection
title_short Dual-Branch Cross-Fusion Normalizing Flow for RGB-D Track Anomaly Detection
title_sort dual branch cross fusion normalizing flow for rgb d track anomaly detection
topic anomaly detection
normalizing flow
RGB-D fusion
url https://www.mdpi.com/1424-8220/25/8/2631
work_keys_str_mv AT xiaoronggao dualbranchcrossfusionnormalizingflowforrgbdtrackanomalydetection
AT pengxuwen dualbranchcrossfusionnormalizingflowforrgbdtrackanomalydetection
AT jinlongli dualbranchcrossfusionnormalizingflowforrgbdtrackanomalydetection
AT linluo dualbranchcrossfusionnormalizingflowforrgbdtrackanomalydetection