Loop Closure Detection Based on Compressed ConvNet Features in Dynamic Environments
Source abstract
Abstract
In dynamic environments, convolutional neural networks (CNNs) often produce image feature maps with significant redundancy due to external factors such as moving objects and occlusions. These feature maps are inadequate as precise image descriptors for similarity measurement, hindering loop closure detection. Addressing this issue, this paper proposes feature compression of convolutional neural network output. The approach is detailed as follows: (1) employing ResNet152 as the backbone feature-extraction network, a Siamese neural network is constructed to enhance the efficiency of feature extraction; (2) utilizing KL transformation to extract principal components from the backbone network’s output, thereby eliminating redundant information; (3) employing the compressed features as input for NetVLAD to construct a spatially informed feature descriptor for similarity measurement. Experimental results demonstrate that, on the New College dataset, the proposed improved method exhibits an approximately 9.98% enhancement in average accuracy compared to the original network. On the City Center dataset, there is an improvement of approximately 2.64%, with an overall increase of about 23.51% in time performance. These findings indicate that the enhanced ResNet152 performs better than the original network in environments with more moving objects and occlusions.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Advanced Image and Video Retrieval Techniques · Advanced Neural Network Applications · Video Surveillance and Tracking Methods
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
College of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, China
Institute of Intelligent Control and Robotics (IICR), Nanjing Forestry University, Nanjing 210037, China
Funding
National Special Research Fund for Non-profit Sector
References (41)
Show source-provided references
1. MonoSLAM: Real-time single camera SLAM10.1109/TPAMI.2007.1049 ↗
2. Past, present and future of simultaneous localization and mapping: Toward the robust-perception age10.1109/TRO.2016.2624754 ↗
3. Chatial, R., and Laumond, J.P. (1985, January 25–28). Position referencing and consistent world modeling for mobile robots. Proceedings of the IEEE International Conference on Robotics and Automation, St. Louis, MO, USA.
4. Research on SLAM Algorithm of Mobile Robot Based on the Fusion of 2D LiDAR and Depth Camera10.1109/ACCESS.2020.3019659 ↗
5. A Survey on Deep Visual Place Recognition10.1109/ACCESS.2021.3054937 ↗
6. iBoW-LCD: An Appearance-based Loop Closure Detection Approach Using Incremental Bags of Binary Words10.1109/LRA.2018.2849609 ↗
7. Distinctive Image Features from Scale-Invariant Keypoints10.1023/B:VISI.0000029664.99615.94 ↗
8. Speeded-Up Robust Features (SURF)10.1016/j.cviu.2007.09.014 ↗
9. Rublee, E., Rabaud, V., Konolige, K., and Bradski, G. (2011, January 6–13). ORB: An efficient alternative to SIFT or SURF. Proceedings of the 2011 International Conference on Computer Vision, Barcelona, Spain.10.1109/ICCV.2011.6126544 ↗
10. Neural feedback text clustering with BiLSTM-CNN-Kmeans10.1109/ACCESS.2018.2873327 ↗
11. Appearance-only SLAM at Large Scale with FAB-MAP 2.010.1177/0278364910385483 ↗
12. Fast and Incremental Method for Loop-Closure Detection Using Bags of Visual Words10.1109/TRO.2008.2004514 ↗
13. Modest-vocabulary loop-closure detection with incremental bag of tracked words10.1016/j.robot.2021.103782 ↗
14. Li, Y., Wei, W., and Zhu, H. (2023). Incremental Bag of Words with Gradient Orientation Histogram for Appearance-Based Loop Closure Detection. Appl. Sci., 13.10.3390/app13116481 ↗
15. Gao, X., and Zhang, T. (2015, January 28–30). Loop closure detection for visual SLAM systems using deep neural networks. Proceedings of the 2015 34th Chinese Control Conference (CCC), Hangzhou, China.10.1109/ChiCC.2015.7260555 ↗
16. Xia, Y., Li, J., Qi, L., Yu, H., and Dong, J. (2017, January 21–23). An evaluation of deep learning in loop closure detection for visual SLAM. Proceedings of the 2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData), Exeter, UK.10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18 ↗
17. Fast and robust loop-closure detection via convolutional auto-encoder and motion consensus10.1109/TII.2021.3120141 ↗
18. Zhang, X., Zheng, L., Tan, Z., and Li, S. (2022). Loop Closure Detection Based on Residual Network and Capsule Network for Mobile Robot. Sensors, 22.10.3390/s22197137 ↗
19. Gradient-based learning applied to document recognition
20. ImageNet Large Scale Visual Recognition Challenge10.1007/s11263-015-0816-y ↗
21. Places: A 10 Million Image Database for Scene Recognition10.1109/TPAMI.2017.2723009 ↗
22. Lin, T.Y., Maire, M., Belongie, S., Bourdev, L., Girshick, R., Hays, J., Perona, P., Zitnick, C.L., and Dollár, P. (2014). Microsoft COCO: Common Objects in Context, Springer International Publishing.10.1007/978-3-319-10602-1_48 ↗
23. Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., Guadarrama, S., and Darrell, T. (2014, January 3–7). Caffe: Convolutional architecture for fast feature embedding. Proceedings of the 22nd ACM international Conference on Multimedia, Orlando, FL, USA.10.1145/2647868.2654889 ↗
24. Hou, Y., Zhang, H., and Zhou, S. (2015, January 8–10). Convolutional neural network-based image representation for visual loop closure detection. Proceedings of the 2015 IEEE International Conference on Information and Automation, Lijiang, China.10.1109/ICInfA.2015.7279659 ↗
25. Sünderhauf, N., Shirazi, S., Dayoub, F., Upcroft, B., and Milford, M. (2015–2, January 28). On the performance of convnet features for place recognition. Proceedings of the 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS), Hamburg, Germany.10.1109/IROS.2015.7353986 ↗
26. Liu, Y., Xiang, R., Zhang, Q., Ren, Z., and Cheng, J. (2019, January 21–23). Loop closure detection based on improved hybrid deep learning architecture. Proceedings of the 2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS), Shenyang, China.10.1109/IUCC/DSCI/SmartCNS.2019.00079 ↗
27. Kim, J.J., Urschler, M., Riddle, P.J., and Wicker, J.S. (2021–1, January 27). SymbioLCD: Ensemble-Based Loop Closure Detection using CNN-Extracted Objects and Visual Bag-of-Words. Proceedings of the 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Prague, Czech Republic.10.1109/IROS51168.2021.9636622 ↗
28. Kuang, X., Guo, J., Bai, J., Geng, H., and Wang, H. (2023). Crop-Planting Area Prediction from Multi-Source Gaofen Satellite Images Using a Novel Deep Learning Model: A Case Study of Yangling District. Remote Sens., 15.10.3390/rs15153792 ↗
29. Fast and robust loop-closure detection using deep neural networks and matrix transformation for a visual SLAM system10.1117/1.JEI.31.6.061816 ↗
30. Hossain, M.M., Hossain, M.A., Musa Miah, A.S., Okuyama, Y., Tomioka, Y., and Shin, J. (2023). Stochastic Neighbor Embedding Feature-Based Hyperspectral Image Classification Using 3D Convolutional Neural Network. Electronics, 12.10.3390/electronics12092082 ↗
31. Signature Verification using a “Siamese” Time Delay Neural Network10.1142/S0218001493000339 ↗
32. Canziani, A., Paszke, A., and Culurciello, E. (2016). An Analysis of Deep Neural Network Models for Practical Applications. arXiv.
33. He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27–30). Deep Residual Learning for Image Recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.10.1109/CVPR.2016.90 ↗
34. Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization10.1007/s11263-019-01228-7 ↗
35. Bannour, S., and Azimi-Sadjadi, M.R. (1991, January 18–21). Principal component extraction using recursive least squares learning method. Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, Singapore.10.1109/IJCNN.1991.170699 ↗
36. Approximating the KLT by Maximizing the Sum of Fourth-Order Moments10.1109/LSP.2013.2237764 ↗
37. NetVLAD: CNN Architecture for Weakly Supervised Place Recognition10.1109/TPAMI.2017.2711011 ↗
38. Milford, M.J., and Wyeth, G.F. (2012, January 14–18). SeqSLAM: Visual route-based navigation for sunny summer days and stormy winter nights. Proceedings of the 2012 IEEE International Conference on Robotics and Automation, Saint Paul, MN, USA.10.1109/ICRA.2012.6224623 ↗
39. Paul, R., and Newman, P. (2010, January 3–7). FAB-MAP 3D: Topological mapping with spatial and visual appearance. Proceedings of the 2010 IEEE International Conference on Robotics and Automation, Anchorage, AK, USA.10.1109/ROBOT.2010.5509587 ↗
40. Simonyan, K., and Zisserman, A. (2014). Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv.
41. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks10.1109/TPAMI.2016.2577031 ↗