This article analyzes the advantages of using multimodal data for disease identification in the detection of rail top surface damage based on deep learning. It comprehensively elaborates on the latest achievements of multimodal deep learning and their application in rail top surface damage detection, while clarifying the challenges and future research directions faced by multimodal disease detection based on deep learning. Firstly, the basic theory and research status of multimodal deep learning are analyzed from perspectives such as multimodal joint and collaborative feature representation, multimodal explicit alignment and implicit alignment, and different methods of multimodal fusion. This paper provides a detailed overview of the latest deep learning-based data fusion algorithms and disease detection methods between different modalities, based on multimodal data such as rail vibration signals, inspection images, and 3D point clouds. The research analysis shows that for multimodal data of steel rails, current research mainly includes three methods related to identification and detection: multimodal representation, multimodal alignment, and multimodal fusion. Current research on multimodal rail top damage detection based on deep learning mainly includes fusion detection of vibration signals and inspection images, as well as fusion detection methods of grayscale image data and 3D point cloud data. Overall, using multimodal deep learning techniques for rail defect identification can effectively enhance accuracy, to a certain extent, eliminate false positives. In the future, learning feature representations based on enhanced semantic sharing and complementarity for multimodal data of steel rails, hybrid alignment models combining feature point alignment and implicit alignment, transformer-based multimodal fusion detection, and missing modal fusion detection will become key research areas for rail top damage detection, providing valuable references for engineering applications.
早期融合方法又称为特征级融合,为了解决各模态中原始数据的不一致问题,可以从每个模态中分别提取特征的表示,然后在特征级别进行融合[45-46]。Fang等[47]提出了一种简单有效的跨模态特征融合方法,利用转换器(Transformer)网络同时执行模态内和模态间的融合,捕获彩色(Red Green Blue,RGB)和热红外领域之间的潜在相互作用,显著提高了多光谱目标检测的性能。Zhu等[48]提出了多模态特征金字塔变换融合2种模态数据,实现不同尺度和不同模态的区域之间的高效关联。模态的相关性在特征层的提取难度很大,Hinton等[49]认为必须经过高层次的抽象特征提取,才能使不同的数据流包含的信息之间存在高相关性。Martínez等[50]指出,多模态数据的早期融合会导致重复信息的出现,加重数据的冗余性,不能充分利用模态间的互补性。
神经网络方法是目前在多模态融合方面应用最广泛的方法之一。深度网络中的不同模态特征向量融合方式主要有拼接、按位乘、按位加、双线性池化、编码器和基于注意力等机制。Fan等[66]将文本特征和音频特征串联送入全连接层对多模态特征进行融合。Yin等[67]使用双向长短期记忆网络(Bi-directional Long Short-Term Memory,BiLSTM)对3种模态进行集成,通过隐藏状态和置信度得分加权得到对应的融合特征。Schneider等[68]提出一种基于编码器的文本到音乐的生成方法,使用文本编码器以及两阶段扩散模型实时生成高质量的立体声音乐。Ji等[69]提出语音-视频联合驱动的3D人脸动画生成方法,编码器生成的语音特征被输入到表情姿势网络中进行特征融合,并使用解码器重建为3D人脸。
Chen等[78]设计了包含多源数据特征提取和多尺度特征融合2个主要网络的多源数据融合算法(Camera and Ultrasound Data Fusion,CUFuse),用于分别提取钢轨图像和超声扫描图像的特征信息并进行融合,对钢轨顶面状态进行分类,其网络模型如图8所示。从图8可以看出:该算法通过双边滤波、索贝尔(Sobel)边缘检测分割轨面区域,多源数据特征提取网络分别提取基于灰度图像和超声图像中轨面状态的信息,多尺度特征融合网络融合多源数据特征提取的不同尺度特征信息;最后,经全连接层预测出钢轨顶面状态的具体类型,即轻、中、重度伤损及正常钢轨母材和接头。
Wang等[80]提出一种钢轨RGB图像与深度图融合的网络(Depth-plus Region Fusion Network,DRFN)解决使用手机拍摄的钢轨图像磨损程度检测,该网络包括深度估计、图像分割和数据融合3个部分,模型如图9所示。从图9可以看出:首先,该网络利用改进的MiDaS深度模型估计的深度图作为指导,充分提取光带的深度信息;其次,基于改进的Mask-RCNN对钢轨光带图像进行分割和提取,改进检测锚框的长宽比例,以适应铁路场景中的目标轨道检测;最后,提出了一种数据融合方法,通过自适应融合模块集成了提取的深度和形状特征,构建了钢轨磨损分类的双通道注意力融合网络,同时加入了其他与钢轨磨损相关的特征向量共同进行钢轨磨损检测。
NIUDaoan, QUJianjun, YANGFei, et al. Application of Track Inspection Big Data of Track Inspection in Track Maintenance and Repair [J]. China Railway, 2022 (11): 8-15. in Chinese
ZHAOHongwei, ZHENGJiajun, ZHAOXinxin, et al. Rail Surface Defect Detection Method Based on BiModal Deep Learning [J]. Computer Engineering and Applications, 2023, 59 (7): 285-293. in Chinese
GUOHao, WANTianyi, YUXiao, et al. Intelligent Detection of Foreign Objects in Railway Catenary Based on Neural Network Fusion Model [J]. Railway Computer Application, 2024, 33 (7): 1-6. in Chinese
[7]
丁要男.钢轨表面缺陷图像分割关键技术研究[D].南昌:华东交通大学,2022.
[8]
DINGYaonan. Research on Key Technologies for Rail Surface Defect Image Segmentation [D]. Nanchang: East China Jiaotong University, 2022. in Chinese
[9]
郭继承.基于深度学习的钢轨表面缺陷自动巡检系统研究[D].兰州:兰州交通大学,2023.
[10]
GUOJicheng. Research on Automatic Inspection System of Rail Surface Defects Based on Deep Learning [D]. Lanzhou: Lanzhou Jiaotong University, 2023. in Chinese
[11]
BAIT B, GAOJ L, YANGJ W, et al. A Study on Railway Surface Defects Detection Based on Machine Vision [J]. Entropy, 2021, 23 (11): 1437.
[12]
ZHENGS B, ZHONGQ W, CHENX Q, et al. The Rail Surface Defects Recognition via Operating Service Rail Vehicle Vibrations [J]. Machines, 2022, 10 (9): 796.
[13]
孙秀秀.基于深度学习的高铁钢轨波磨检测的研究[D].北京:北京建筑大学,2020.
[14]
SUNXiuxiu. Research on High-Speed Rail Corrugation Detection Based on Deep Learning [D]. Beijing: Beijing University of Civil Engineering and Architecture, 2020. in Chinese
[15]
黄振宇.深度多模态表示学习研究[D].成都:四川大学,2023.
[16]
HUANGZhenyu. Research on Deep Multimodal Representation Learning [D]. Chengdu: Sichuan University, 2023. in Chinese
[17]
GUOW Z, WANGJ W, WANGS P. Deep Multimodal Representation Learning: a Survey [J]. IEEE Access, 2019, 7: 63373-63394.
[18]
FANW Q, HEZ W, XINGX F, et al. Multi-Modality Depression Detection via Multi-Scale Temporal Dilated CNNS [C]// Proceedings of the 9th International on Audio/Visual Emotion Challenge and Workshop. Nice, France, New York: ACM, 2019: 73-80.
[19]
YINS, LIANGC, DINGH Y, et al. A Multi-Modal Hierarchical Recurrent Neural Network for Depression Detection [C]// Proceedings of the 9th International on Audio/Visual Emotion Challenge and Workshop. Nice, France. New York: ACM, 2019: 65-71.
[20]
CHENGX L, GENGK K, WANGZ W, et al. SLBAF-Net: Super-Lightweight Bimodal Adaptive Fusion Network for UAV Detection in Low Recognition Environment [J]. Multimedia Tools and Applications, 2023, 82 (30): 47773-47792.
[21]
LIUM H, SHIR X, KUANGK M. OpenShape: Scaling up 3D Shape Representation towards Open-World Understanding [C]// NIPS '23: Proceedings of the 37th International Conference on Neural Information Processing Systems. New Orleans, LA, USA. Red Hook: Curran Associates Inc., 2024: 44860-44879.
[22]
KIMY, LEEH, PROVOSTE M. Deep Learning for Robust Feature Generation in Audiovisual Emotion Recognition [C]// Proceedings of the 2013 IEEE International Conference on Acoustics, Speech and Signal Processing. Vancouver, BC, Canada. New York: IEEE Press, 2013: 3687-3691.
[23]
TANGC, YANGX, WUB J B, et al. Parts2Words: Learning Joint Embedding of Point Clouds and Texts by Bidirectional Matching between Parts and Words [C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver, BC, Canada. New York: IEEE, 2023: 6884-6893.
[24]
KAHOUS E, BOUTHILLIERX, LAMBLINP, et al. EmoNets: Multimodal Deep Learning Approaches for Emotion Recognition in Video [J]. Journal on Multimodal User Interfaces, 2016, 10 (2): 99-111.
[25]
NGIAMJ, KHOSLAA, KIMM, et al. Multimodal Deep Learning[C]// Proceedings of the 28th International Conference on Machine Learning (ICML-11). Bellevue, Washington, USA. Madison, WI: Omnipress, 2011: 689-696.
[26]
SRIVASTAVAN, SALAKHUTDINOVR. Multimodal Learning with Deep Boltzmann Machines [J]. The Journal of Machine Learning Research, 2014, 15 (1): 2949-2980.
[27]
WANGS P, ZHANGH D, WANGH. Object Co-Segmentation via Weakly Supervised Data Fusion [J]. Computer Vision and Image Understanding, 2017, 155: 43-54.
[28]
NGIAMJ, KHOSLAA, KIMM, et al. Multimodal Deep Learning [C]// Proceedings of the 28th International Conference on Machine Learning (ICML-11). Bellevue, Washington, USA. Madison, WI: Omnipress, 2011: 689-696.
[29]
JIANGY G, WUZ X, WANGJ, et al. Exploiting Feature and Class Relationships in Video Categorization with Regularized Deep Neural Networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40 (2): 352-364.
CAOY, LONGM S, WANGJ M, et al. Deep Visual-Semantic Hashing for Cross-Modal Retrieval [C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco, California, USA. New York: ACM, 2016: 1445-1454.
[32]
WESTONJ, BENGIOS, USUNIERN. Large Scale Image Annotation: Learning to Rank with Joint Word-Image Embeddings [J]. Machine Learning, 2010, 81 (1): 21-35.
[33]
FROMEA, CORRADOG S, SHLENSJ, et al. DeViSE [C]// Proceedings of the 27th International Conference on Neural Information Processing Systems - Volume 2. Lake Tahoe, Nevada. New York: ACM, 2013: 2121-2129.
[34]
REEDS, AKATAZ, LEEH, et al. Learning Deep Representations of Fine-Grained Visual Descriptions [C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV, USA. New York: IEEE Press, 2016: 49-58.
[35]
LIUD N, ZHANGD H, SONGY, et al. PDAM: a Panoptic-Level Feature Alignment Framework for Unsupervised Domain Adaptive Instance Segmentation in Microscopy Images [J]. IEEE Transactions on Medical Imaging, 2021, 40 (1): 154-165.
[36]
CHENB Z, CAOQ, HOUM X, et al. Multimodal Emotion Recognition with Temporal and Semantic Consistency [J]. IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2021, 29: 3592-3603.
[37]
LIK S, CHENC, CAOW T, et al. DeAF: a Multimodal Deep Learning Framework for Disease Prediction [J]. Computers in Biology and Medicine, 2023, 156: 106715.
[38]
KIMT, CHA M, KIMH, et al. Learning to Discover Cross-Domain Relations with Generative Adversarial Networks [C]// Proceedings of the 34th International Conference on Machine Learning - Volume 70. Sydney, NSW, Australia. JMLR, 2017: 1857-1865.
[39]
CHUNGY A, WENGW H, TONGS, et al. Unsupervised Cross-Modal Alignment of Speech and Text Embedding Spaces [C]// Proceedings of the 32nd International Conference on Neural Information Processing Systems. Montréal, Canada. Red Hook: Curran Associates Inc., 2018: 7365-7375.
[40]
ZHUY K, KIROSR, ZEMELR, et al. Aligning Books and Movies: towards Story-Like Visual Explanations by Watching Movies and Reading Books [C]// 2015 Proceedings of the IEEE International Conference on Computer Vision (ICCV). Santiago, Chile. New York: IEEE Press, 2015: 19-27.
[41]
GEBRUI D, BAS, LIX F, et al. Audio-Visual Speaker Diarization Based on Spatiotemporal Bayesian Fusion [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40 (5): 1086-1099.
[42]
VOGELS, NEYH, TILLMANNC, et al. HMM-Based Word Alignment in Statistical Translation [C]// Proceedings of the 16th Conference on Computational Linguistics - Volume 2. Copenhagen, Denmark. Stroudsburg: Association for Computational Linguistics, 1996: 836-841.
[43]
MESSINAN, AMATOG, ESULIA, et al. Fine-Grained Visual Textual Alignment for Cross-Modal Retrieval Using Transformer Encoders [J]. ACM Transactions on Multimedia Computing, Communications, and Applications, 2021, 17 (4): 1-23.
[44]
XUEF, LIY, LIUD Y, et al. LipFormer: Learning to Lipread Unseen Speakers Based on Visual-Landmark Transformers [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33 (9): 4507-4517.
[45]
QUL G, LIUM, CAOD, et al. Context-Aware Multi-View Summarization Network for Image-Text Matching [C]// Proceedings of the 28th ACM International Conference on Multimedia. Seattle, WA, USA. New York: ACM, 2020: 1047-1055.
[46]
FANGQ Y, HAND P, WANGZ K. Cross-Modality Fusion Transformer for Multispectral Object Detection [EB/OL]. 2021: 2111.00273.
ZHANGHucheng, LILeixiao, LIUDongjiang. Survey of Multimodal Data Fusion Research [J]. Journal of Frontiers of Computer Science and Technology, 2024, 18 (10): 2501-2520. in Chinese
[49]
TEEPET, WOLTERSP, GILGJ, et al. EarlyBird: Early-Fusion for Multi-View Tracking in the Bird's Eye View [C]// 2024 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW). Waikoloa, HI, USA. New York: IEEE Press, 2024: 102-111.
[50]
DAS R, SINGHT D. Image-Text Multimodal Sentiment Analysis Framework of Assamese News Articles Using Late Fusion [J]. ACM Transactions on Asian and Low-Resource Language Information Processing, 2023, 22 (6): 1-30.
[51]
KUMARP, MALIKS, RAMANB. Interpretable Multimodal Emotion Recognition Using Hybrid Fusion of Speech and Image Data [J]. Multimedia Tools and Applications, 2024, 83 (10): 28373-28394.
[52]
ATREYP K, HOSSAINM A, SADDIK AEL, et al. Multimodal Fusion for Multimedia Analysis: a Survey [J]. Multimedia Systems, 2010, 16 (6): 345-379.
[53]
XUEY J, JUZ Y, LIY M, et al. MAF-YOLO: Multi-Modal Attention Fusion Based YOLO for Pedestrian Detection [J]. Infrared Physics & Technology, 2021, 118: 103906.
[54]
ZHAOZ X, BAIH, ZHANGJ S, et al. Cddfuse: Correlation-Driven Dual-Branch Feature Decomposition for Multi-Modality Image Fusion [C]// 2023 Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver, BC, Canada. New York: IEEE, 2023: 5906-5916.
[55]
FANGQ Y, HAND P, WANGZ K. Cross-Modality Fusion Transformer for Multispectral Object Detection [EB/OL]. 2021: 2111.00273.
[56]
ZHUY H, SUNX Y, WANGM, et al. Multi-Modal Feature Pyramid Transformer for RGB-Infrared Object Detection [J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24 (9): 9984-9995.
[57]
HINTONG E, SALAKHUTDINOVR R. Reducing the Dimensionality of Data with Neural Networks [J]. Science, 2006, 313 (5786): 504-507.
[58]
MARTÍNEZH P, YANNAKAKISG N. Deep Multimodal Fusion: Combining Discrete Events and Continuous Signals [C]// Proceedings of the 16th International Conference on Multimodal Interaction. Istanbul Turkey. New York: ACM, 2014: 34-41.
[59]
MENGH Y, HUANGD, WANGH, et al. Depression Recognition Based on Dynamic Facial and Vocal Expression Features Using Partial Least Square Regression [C]// Proceedings of the 3rd ACM International Workshop on Audio/Visual Emotion Challenge. Barcelona, Spain. New York: ACM, 2013: 21-30.
[60]
SAMAREHA, JINY, WANGZ Y, et al. Detect Depression from Communication: How Computer Vision, Signal Processing, and Sentiment Analysis Join Forces [J]. IISE Transactions on Healthcare Systems Engineering, 2018, 8 (3): 196-208.
[61]
ZHANGQ J, ZHANGH T, ZHOUK M, et al. Developing a Physiological Signal-Based, Mean Threshold and Decision-Level Fusion Algorithm (PMD) for Emotion Recognition [J]. Tsinghua Science and Technology, 2023, 28 (4): 673-685.
[62]
MEHRAS, RANGAV, AGARWALR. Improving Speech Command Recognition through Decision-Level Fusion of Deep Filtered Speech Cues [J]. Signal, Image and Video Processing, 2024, 18 (2): 1365-1373.
[63]
LANZ Z, BAOL, YUS I, et al. Multimedia Classification and Event Detection Using Double Fusion [J]. Multimedia Tools and Applications, 2014, 71 (1): 333-347.
[64]
ALGHOWINEMS, GOECKER, COHNJ F, et al. Cross-Cultural Detection of Depression from Nonverbal Behaviour [C]// 2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG). Ljubljana, Slovenia. New York: IEEE Press, 2015: 1-8.
[65]
MORALESM, SCHERERS, LEVITANR. A Linguistically-Informed Fusion Approach for Multimodal Depression Detection [C]// Proceedings of the Fifth Workshop on Computational Linguistics and Clinical Psychology: from Keyboard to Clinic. New Orleans, LA. Stroudsburg: USAACL, 2018: 13-24.
[66]
WANGY K, CHENX H, CAOL L, et al. Multimodal Token Fusion for Vision Transformers [C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, LA, USA. New York: IEEE Press, 2022: 12176-12185.
[67]
BRANDERC, CIOFLANC, NICULESCUV, et al. Improving Data-Scarce Image Classification through Multimodal Synthetic Data Pretraining [C]// 2023 IEEE Sensors Applications Symposium (SAS). Ottawa, ON, Canada. New York: IEEE Press, 2023: 1-6.
[68]
JIANGX Y, WUF, ZHANGY, et al. The Classification of Multi-Modal Data with Hidden Conditional Random Field [J]. Pattern Recognition Letters, 2015, 51: 63-69.
[69]
WÖLLMERM, METALLINOUA, EYBENF, et al. Context-Sensitive Multimodal Emotion Recognition from Speech and Facial Expression Using Bidirectional LSTM Modeling [C]// 11th Annual Conference of the International Speech Communication Association, Interspeech. Makuhari, Chiba, Japan. Illinois: ISCA, 2010: 2362-2365.
[70]
SIKKAK, DYKSTRAK, SATHYANARAYANAS, et al. Multiple Kernel Learning for Emotion Recognition in the Wild [C]// Proceedings of the 15th ACM on International Conference on Multimodal Interaction. Sydney, Australia. New York: ACM, 2013: 517-524.
[71]
JAQUESN, TAYLORS, SANOA, et al. Multi-Task, Multi-Kernel Learning for Estimating Individual Wellbeing [C]// Proceedings of NIPS Workshop on Multimodal Machine Learning, Montreal, Quebec. San Diego: NeurIPS, 2015, 898: 3.
[72]
SUTTONC, MCCALLUMA. Introduction to Statistical Relational Learning [M]. Cambridge: The MIT Press, 2007: 93-128.
[73]
AFOURAST, CHUNGJ S, SENIORA, et al. Deep Audio-Visual Speech Recognition [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44 (12): 8717-8727.
[74]
FANW Q, HEZ W, XINGX F, et al. Multi-Modality Depression Detection via Multi-Scale Temporal Dilated CNNS [C]// Proceedings of the 9th International on Audio/Visual Emotion Challenge and Workshop. Nice, France. New York: ACM, 2019: 73-80.
[75]
YINS, LIANGC, DINGH Y, et al. A Multi-Modal Hierarchical Recurrent Neural Network for Depression Detection [C]// Proceedings of the 9th International on Audio/Visual Emotion Challenge and Workshop. Nice, France. New York: ACM, 2019: 65-71.
[76]
SCHNEIDERF, KAMALO, JINZ J, et al. Moûsai: Text-to-Music Generation with Long-Context Latent Diffusion [EB/OL]. 2023: 2301.11757.
[77]
JIX J, LIAOZ Z, DONGL F, et al. 3D Facial Animation Driven by Speech-Video Dual-Modal Signals [J]. Complex & Intelligent Systems, 2024, 10 (5): 5951-5964.
[78]
JAEGLEA, GIMENOF, BROCKA, et al. Perceiver: General Perception with Iterative Attention [C]// International Conference on Machine Learning. Virtual. PMLR, 2021: 4651-4664.
[79]
SUNC, MYERSA, VONDRICKC, et al. VideoBERT: a Joint Model for Video and Language Representation Learning [C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Seoul, South Korea. PMLR, 2019: 7464-7473.
[80]
HUANGJ, TAOJ H, LIUB, et al. Multimodal Transformer Fusion for Continuous Emotion Recognition [C]// ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Barcelona, Spain. New York: IEEE Press, 2020: 3507-3511.
[81]
ZHANGW, QIUF, WANGS, et al. Transformer-Based Multimodal Information Fusion for Facial Expression Analysis [C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver, BC, Canada. New York: 2022: 2428-2437.
[82]
ALABAS Y, BALLJ E. Transformer-Based Optimized Multimodal Fusion for 3D Object Detection in Autonomous Driving [J]. IEEE Access, 2024, 12: 50165-50176.
[83]
ZHOUY, YANGC, WANGP, et al. ViT-FuseNet: Multimodal Fusion of Vision Transformer for Vehicle-Infrastructure Cooperative Perception [J]. IEEE Access, 2024, 12: 31640-31651.
[84]
YANGT T, XUT H, CHENGY, et al. A Fusion Method Based on 1D Vibration Signals and 2D Images for Detection of Railway Surface Defects [C]// 2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE). Guangzhou, China. New York: IEEE Press, 2023: 282-286.
[85]
SHENY F, ZHONGQ W, ZHENGS B, et al. A Multi-Modal Approach to Rail Surface Condition Analysis: the MFDF-Net [J]. IEEE Access, 2024, 12: 132480-132494.
[86]
CHENZ X, WANGQ H, HEQ, et al. CUFuse: Camera and Ultrasound Data Fusion for Rail Defect Detection [J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23 (11): 21971-21983.
[87]
WUJ Y, ZHOUW J, QIUW W, et al. Depth Repeated-Enhancement RGB Network for Rail Surface Defect Inspection [J]. IEEE Signal Processing Letters, 2022, 29: 2053-2057.
[88]
WANGQ H, GAOT C, HEQ, et al. Severe Rail Wear Detection with Rail Running Band Images [J]. Computer-Aided Civil and Infrastructure Engineering, 2023, 38 (9): 1162-1180.
[89]
ZHOUW J, HONGJ K. FHENet: Lightweight Feature Hierarchical Exploration Network for Real-Time Rail Surface Defect Inspection in RGB-D Images [J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 5005008.
[90]
ZHANGJ X, ZHANGJ, CHENJ J, et al. Rail Surface Defect Detection through Bimodal RSDINet and Three-Branched Evidential Fusion [J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 2508714.
[91]
YANGJ X, ZHOUW J, WUR M, et al. CSANet: Contour and Semantic Feature Alignment Fusion Network for Rail Surface Defect Detection [J]. IEEE Signal Processing Letters, 2023, 30: 972-976.
[92]
WANGJ, LIG Q, QIUG W, et al. Depth-Assisted Semi-Supervised RGB-D Rail Surface Defect Inspection [J]. IEEE Transactions on Intelligent Transportation Systems, 2024, 25 (7): 8042-8052.
HEQing, MAYusong, LIChenzhong, et al. High-Speed Railway Mileage Alignment Matching and Error Correction between Dynamic and Static Track Inspection Data [J]. Journal of the China Railway Society, 2024, 46 (1): 129-136. in Chinese