Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation
Deep neural networks are effectively utilized for the instance segmentation of muck images from tunnel boring machines (TBMs), providing real-time insights into the surrounding rock condition. However, the high cost of obtaining quality labeled data limits the widespread application of this method....
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Symmetry |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-8994/16/2/222 |
| _version_ | 1797296959238701056 |
|---|---|
| author | Ke Lei Zhongsheng Tan Xiuying Wang Zhenliang Zhou |
| author_facet | Ke Lei Zhongsheng Tan Xiuying Wang Zhenliang Zhou |
| author_sort | Ke Lei |
| collection | DOAJ |
| description | Deep neural networks are effectively utilized for the instance segmentation of muck images from tunnel boring machines (TBMs), providing real-time insights into the surrounding rock condition. However, the high cost of obtaining quality labeled data limits the widespread application of this method. Addressing this challenge, this study presents a semi-symmetrical, fully convolutional masked autoencoder designed for self-supervised pre-training on extensive unlabeled muck image datasets. The model features a four-tier sparse encoder for down-sampling and a two-tier sparse decoder for up-sampling, connected via a conventional convolutional neck, forming a semi-symmetrical structure. This design enhances the model’s ability to capture essential low-level features, including geometric shapes and object boundaries. Additionally, to circumvent the trivial solutions in pixel regression that the original masked autoencoder faced, Histogram of Oriented Gradients (HOG) descriptors and Laplacian features have been integrated as novel self-supervision targets. Testing shows that the proposed model can effectively discern essential features of muck images in self-supervised training. When applied to subsequent end-to-end training tasks, it enhances the model’s performance, increasing the prediction accuracy of Intersection over Union (IoU) for muck boundaries and regions by 5.9% and 2.4%, respectively, outperforming the enhancements made by the original masked autoencoder. |
| first_indexed | 2024-03-07T22:12:22Z |
| format | Article |
| id | doaj.art-bb47ebe322624061b9c25142d83a6623 |
| institution | Directory Open Access Journal |
| issn | 2073-8994 |
| language | English |
| last_indexed | 2024-03-07T22:12:22Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Symmetry |
| spelling | doaj.art-bb47ebe322624061b9c25142d83a66232024-02-23T15:36:03ZengMDPI AGSymmetry2073-89942024-02-0116222210.3390/sym16020222Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image SegmentationKe Lei0Zhongsheng Tan1Xiuying Wang2Zhenliang Zhou3Key Laboratory of Urban Underground Engineering of Ministry of Education, Beijing Jiaotong University, Beijing 100044, ChinaKey Laboratory of Urban Underground Engineering of Ministry of Education, Beijing Jiaotong University, Beijing 100044, ChinaKey Laboratory of Urban Underground Engineering of Ministry of Education, Beijing Jiaotong University, Beijing 100044, ChinaKey Laboratory of Urban Underground Engineering of Ministry of Education, Beijing Jiaotong University, Beijing 100044, ChinaDeep neural networks are effectively utilized for the instance segmentation of muck images from tunnel boring machines (TBMs), providing real-time insights into the surrounding rock condition. However, the high cost of obtaining quality labeled data limits the widespread application of this method. Addressing this challenge, this study presents a semi-symmetrical, fully convolutional masked autoencoder designed for self-supervised pre-training on extensive unlabeled muck image datasets. The model features a four-tier sparse encoder for down-sampling and a two-tier sparse decoder for up-sampling, connected via a conventional convolutional neck, forming a semi-symmetrical structure. This design enhances the model’s ability to capture essential low-level features, including geometric shapes and object boundaries. Additionally, to circumvent the trivial solutions in pixel regression that the original masked autoencoder faced, Histogram of Oriented Gradients (HOG) descriptors and Laplacian features have been integrated as novel self-supervision targets. Testing shows that the proposed model can effectively discern essential features of muck images in self-supervised training. When applied to subsequent end-to-end training tasks, it enhances the model’s performance, increasing the prediction accuracy of Intersection over Union (IoU) for muck boundaries and regions by 5.9% and 2.4%, respectively, outperforming the enhancements made by the original masked autoencoder.https://www.mdpi.com/2073-8994/16/2/222intelligent TBM tunnelingreal-time muck analysisself-supervised traininginstance segmentationfully convolutional masked autoencoderHOG descriptor |
| spellingShingle | Ke Lei Zhongsheng Tan Xiuying Wang Zhenliang Zhou Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation Symmetry intelligent TBM tunneling real-time muck analysis self-supervised training instance segmentation fully convolutional masked autoencoder HOG descriptor |
| title | Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation |
| title_full | Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation |
| title_fullStr | Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation |
| title_full_unstemmed | Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation |
| title_short | Semi-Symmetrical, Fully Convolutional Masked Autoencoder for TBM Muck Image Segmentation |
| title_sort | semi symmetrical fully convolutional masked autoencoder for tbm muck image segmentation |
| topic | intelligent TBM tunneling real-time muck analysis self-supervised training instance segmentation fully convolutional masked autoencoder HOG descriptor |
| url | https://www.mdpi.com/2073-8994/16/2/222 |
| work_keys_str_mv | AT kelei semisymmetricalfullyconvolutionalmaskedautoencoderfortbmmuckimagesegmentation AT zhongshengtan semisymmetricalfullyconvolutionalmaskedautoencoderfortbmmuckimagesegmentation AT xiuyingwang semisymmetricalfullyconvolutionalmaskedautoencoderfortbmmuckimagesegmentation AT zhenliangzhou semisymmetricalfullyconvolutionalmaskedautoencoderfortbmmuckimagesegmentation |