Abstract
With rapid developments in the fields of autonomous driving, robot navigation, and augmented reality, visual SLAM technology has become one of the core key technologies. While VSLAM systems perform more consistently in static scenes, introducing dynamic objects such as people, vehicles, and animals into the scene makes reliable map building and localization more difficult, and accurate trajectory estimation more challenging to achieve. In this paper, we propose a semantic VSLAM system based on the Global attention mechanism (GAM) and adaptive thresholding. First, GAM improves the segmentation accuracy of the Mask R-CNN network model for dynamic objects and eliminates the influence of dynamic objects on the VSLAM system. In addition, adaptive thresholding generates adaptive factors based on the number of key points extracted in the scene and dynamically adjusts the FAST threshold, which enables more stable extraction of feature points in dynamic scenes. We have verified our approach on the TUM public dataset, and compared with the DynaSLAM method. The absolute trajectory error (ATE) and relative trajectory error (RPE) are reduced to some extent on its dataset. Especially on the W_rpy dataset, the accuracy of our method is improved by 38.78%. The experimental results show that our proposed method can significantly improve the overall performance of the system in highly dynamic environments.
This work was supported in part by the Key Research and Development Project of Hainan Province (ZDYF2022GXJS348, ZDYF2022SHFZ039), the Hainan Province Natural Science Foundation (623RC446) and the National Natural Science Foun- dation of China (62161010, 61963012). The authors would like to thank the referees for their constructive suggestions.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Chang ,Y., Tian, Y., How, J.P., Carlone, L.: Kimera-multi: a system for distributed multi-robot metric-semantic simultaneous localization and mapping. In 2021 IEEE International Conference on Robotics and Automation (ICRA), pp. 11210−11218 (IEEE)
Cheng, J., Zhang, L., Chen, Q., Hu, X., Cai, J.: A review of visual slam methods for autonomous driving vehicles. Appl. Artif. Intell. 114, 104992 (2022)
Jinyu, L., Bangbang, Y., Danpeng, C., Nan, W., Guofeng, Z., Hujun B.: Survey and evaluation of monocular visual-inertial slam algorithms for augmented reality. Virt. Real. Intell. Hardw. 1(4), 386 410 (2019)
Liu, Y., Miura J.: Rds-slam: Real-time dynamic slam using semantic segmentation methods. IEEE Access 9, 23772−23785 (2021)
Wang, H., Ko, J.Y., Xie, L.: Multi-modal semantic slam for complex dynamic environments (2022)
Li, A., Wang, J., Xu, M., Chen, Z.: DP-SLAM: A visual slam with moving probability towards dynamic environments. Science 556, 128−142 (2021)
Mur-Artal, R., Montiel, J.M., Tardos, J.D.: ORB-SLAM: a versatile and accurate monocular SLAM system. IEEE Trans. Robot. 31(5), 1147−1163 (2015)
Mur-Artal, R., Tardós, J.D.: Orb-slam2: an open-source slam system for monocular, stereo, and RGB-D cameras. IEEE Trans. Robot. 33(5), 1255−1262 (2017)
Yu, C., et al.: DS- SLAM: a semantic visual slam towards dynamic environments. In: 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 1168−1174. IEEE (2018)
Zhao, C., Sun, L., Purkait, P., Duckett, T., Stolkin, R.: Learning monocular visual odometry with dense 3d mapping from dense 3d ow. In: 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 6864 6871. IEEE (2018)
Konda, K.R., Memisevic, R.: Learning visual odometry with a convolutional network. VISAPP 486(490), 2015 (2015)
Wu, W., Guo, L., Gao, H., You, Z., Liu, Y., Chen, Z.: Yolo-slam: a semantic slam system to- wards dynamic environment with geometric constraint. Neural Comput. Appl. 1−16 (2022)
Redmon, J., Farhadi, A.: Yolov3: An incremental improvement (2018)
Bescos, B., Fácil, J.M., Civera, J., Neira, J.: DynaSLAM: Tracking, mapping, and inpainting in dynamic scenes. IEEE Robot. Autom. Lett. 3(4), 4076−4083 (2018)
Alismail, H., Kaess, M., Browning, B., Lucey, S.: Direct visual odometry in low light using binary descriptors. IEEE Robot. Autom. Lett. 2(2), 444−451(2016)
Ono, Y., Trulls, E., Fua, P., Yi, K.M.: Learning local features from images. Inf. Process. Syst. Yi. Lf-net. 31 (2018)
Liu, Y., Shao, Z., Hoffmann, N.: Global attention mechanism: retain information to enhance channel-spatial interactions (2021)
He, K., Gkioxari, G., Dollár, P., Girshick, R.: Mask r-cnn. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2961−2969 (2017)
Woo, S., Park, J., Lee, J.Y., Kweon, I.S.: Cbam: convolutional block attention module. In Proceedings of the European Conference on Computer Vision (ECCV), pp. 3−19 (2318)
Girshick, R.: Fast r-cnn. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 1440−1448 (2015)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770−778 (2016)
Lin, T.-Y., et al.: Microsoft COCO: common objects in context. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8693, pp. 740–755. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-10602-1_48
Sturm, J., Engelhard, N., Endres, F., Burgard, W., Cremers, D.: A benchmark for the evaluation of RGB-D SLAM systems. In: 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 573 580. IEEE (2012)
Acknowledgements
We thank Shenzhen Umouse Technology Development Co., Ltd. For their support in equipments and experimental conditions.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Cai, D., Hu, Z., Li, R., Qi, H., Xiang, Y., Zhao, Y. (2023). AGAM-SLAM: An Adaptive Dynamic Scene Semantic SLAM Method Based on GAM. In: Huang, DS., Premaratne, P., Jin, B., Qu, B., Jo, KH., Hussain, A. (eds) Advanced Intelligent Computing Technology and Applications. ICIC 2023. Lecture Notes in Computer Science(), vol 14090. Springer, Singapore. https://doi.org/10.1007/978-981-99-4761-4_3
Download citation
DOI: https://doi.org/10.1007/978-981-99-4761-4_3
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-4760-7
Online ISBN: 978-981-99-4761-4
eBook Packages: Computer ScienceComputer Science (R0)