The DOI matches its Crossref record; 125 of the 127 references it cites were found in a registry or an index.
Approaches, Challenges, and Applications for Deep Visual Odometry: Toward Complicated and Emerging Areas
Open accessAbstract
Visual odometry (VO) is a prevalent way to deal with the relative localization problem, which is becoming increasingly mature and accurate, but it tends to be fragile under challenging environments. Comparing with classical geometry-based methods, deep-learning-based methods can automatically learn effective and robust representations, such as depth, optical flow, feature, ego-motion, etc., from data without explicit computation. Nevertheless, there still lacks a thorough review of the recent advances of deep-learning-based VO (Deep VO). Therefore, this article aims to gain a deep insight on how deep learning can profit and optimize the VO systems. We first screen out a number of qualifications, including accuracy, efficiency, scalability, dynamicity, practicability, and extensibility, and employ them as the criteria. Then, using the offered criteria as the uniform measurements, we detailedly evaluate and discuss how deep learning improves the performance of VO from the aspects of depth estimation, feature extraction and matching, and pose estimation. We also summarize the complicated and emerging areas of Deep VO, such as mobile robots, medical robots, augmented and virtual reality, etc. Through the literature decomposition, analysis, and comparison, we finally put forward a number of open issues and raise some future research directions in this field.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Are we ready for autonomous driving? The KITTI vision benchmark suite
Geiger, A., Lenz, P., & Urtasun, R. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248074 · Cited by 14,897 · cited 1×
“For the former, we use RMSE (Root Mean Square Error) [23] to measure the bias of ground truth and estimation, which directly describes algorithm accuracy.”
MethodORB-SLAM: A Versatile and Accurate Monocular SLAM System
Mur-Artal, R., Montiel, J. M. M., & Tardos, J. D. (2015) · IEEE Transactions on Robotics · doi:10.1109/tro.2015.2463671 · Cited by 6,778 · cited 2×
“use root-mean-square error (RMSE, defined in [28]) or average trajectory error (ATE, defined in [4]) to measure the bias of ground truth and estimation, which directly describes algorithm accuracy.”
BackgroundMethodLSD-SLAM: Large-scale direct monocular SLAM
Engel, J., Schöps, T., & Cremers, D. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10605-2_54 · Cited by 3,851
Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age
Cadena, C., Carlone, L., Carrillo, H., Latif, Y., Scaramuzza, D., Neira, J., Reid, I., & Leonard, J. J. (2016) · IEEE Transactions on Robotics · doi:10.1109/tro.2016.2624754 · Cited by 3,596 · cited 1×
“Most recent review papers focus on geometrybased methods, less on deep-learning-based approaches [16], [18], [19].”
BackgroundDepth Map Prediction from a Single Image using a Multi-Scale Deep Network
Eigen, D., Puhrsch, C., & Fergus, R. (2014) · arXiv · doi:10.48550/arxiv.1406.2283 · Cited by 3,442 · cited 1×
“1) Supervised Learning Methods: For supervised learning, the use of supervised-based methods was first proposed in [29].”
MethodUnsupervised Monocular Depth Estimation with Left-Right Consistency
Godard, C., Aodha, O. M., & Brostow, G. J. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.699 · Cited by 3,270 · cited 2×
“[46] introduced a method that can simultaneously infer disparities that warp the left images to match the right one or warp the right image to match the left one, thereby obtaining better depth estimation by enforcing left–right depth consistency.”
BackgroundMethodEngel, J. J., Koltun, V., & Cremers, D. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2658577 · Cited by 2,995
Unsupervised Learning of Depth and Ego-Motion from Video
Zhou, T., Brown, M., Snavely, N., & Lowe, D. G. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.700 · Cited by 2,910 · cited 1×
“[47] first synthesized a new image of the scene from different views of the input image.”
BackgroundPoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization
Kendall, A., Grimes, M., & Cipolla, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.336 · Cited by 2,489 · cited 1×
“1) Supervised Methods: PoseNet [83] was the work that the first to utilize CNN to regress 6 DOF camera pose from the monocular image in an end-to-end manner without graph optimization.”
MethodUnsupervised CNN for Single View Depth Estimation: Geometry to the Rescue
Garg, R., B.G., V. K., Carneiro, G., & Reid, I. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46484-8_45 · Cited by 1,632 · cited 1×
“The method using unsupervised learning was first introduced in [45], which estimates the depth information from single-view images.”
Method
Research works printed but not linked: 13
- Campos, 2020, ORB-SLAM3: An accurate open-source library for visual, visual-inertial and multi-map SLAM
- 10.1007/978-3-030-01237-3_40 · DOI
- Ma, 2020, Progress in research on monocular visual odometry of autonomous vehicles, J. Jilin Univ., 50, 765
- 10.1145/3177853 · DOI
- Zhang, 2019, DFineNet: Ego-motion estimation and depth refinement from sparse, noisy depth input with RGB guidance
- Ramirez, 2018, Geometry meets semantics for semi-supervised monocular depth estimation
- Tosi, 2019, Learning monocular depth estimation infusing traditional stereo knowledge, 10.1109/CVPR.2019.01003 · DOI
- Mishchuk, 2017, Working hard to know your neighbor’s margins: Local descriptor learning loss, 30
- Cieslewski, 2018, Matching features without descriptors: Implicitly matched interest points
- Jiao, 2018, MagicVO: End-to-end monocular visual odometry through deep bi-directional recurrent convolutional neural network
- 10.1007/978-3-030-29911-8_35 · DOI
- Vijayanarasimhan, 2017, SfM-Net: Learning of structure and motion from video
- Wei, 2019, Instant motion tracking and its applications to augmented reality
Cited by
Sorted by citations
Vision-based navigation and guidance for agricultural autonomous vehicles and robots: A review
Bai, Y., Zhang, B., Xu, N., Zhou, J. C., Shi, J., & Diao, Z. (2022) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2022.107584 · Cited by 372
An Overview on Visual SLAM: From Tradition to Semantic
Chen, W.-F., Shang, G., Ji, A.-H., Zhou, C., Wang, X., Xu, C., Li, Z., & Hu, K. (2022) · Remote Sensing · doi:10.3390/rs14133010 · Cited by 217
Performance and Challenges of 3D Object Detection Methods in Complex Scenes for Autonomous Driving
Wang, K., Zhou, T., Li, X., & Ren, F. (2022) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2022.3213796 · Cited by 143
“R OBUST and accurate 3D object detection methods are essential for AD vehicles to understand the surrounding environment [1].”
MethodSelf-citationYu, S., Liu, X., Tan, Q., Wang, Z., & Zhang, B. (2024) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2024.109229 · Cited by 103
Deep Learning Techniques for Visual SLAM: A Survey
Mokssit, S., Licea, D. B., Guermah, B., & Ghogho, M. (2023) · IEEE Access · doi:10.1109/access.2023.3249661 · Cited by 81
“[32], [33], visual odometry estimation [34], or loop closure detection [22].”
BackgroundAn Overview of Key SLAM Technologies for Underwater Scenes
Wang, X., Fan, X., Shi, P., Ni, J., & Zhou, Z. (2023) · Remote Sensing · doi:10.3390/rs15102496 · Cited by 70
A Practical Survey on Visual Odometry for Autonomous Driving in Challenging Scenarios and Conditions
Agostinho, L. R., Ricardo, N. M., Pereira, M. I., Hiolle, A., & Pinto, A. M. (2022) · IEEE Access · doi:10.1109/access.2022.3188990 · Cited by 57
Aslan, M. F., Durdu, A., Yusefi, A., & Yılmaz, A. (2022) · Neural Networks · doi:10.1016/j.neunet.2022.09.001 · Cited by 58
Robust visual-based localization and mapping for underwater vehicles: A survey
Ding, S.-S., Zhang, T.-D., Lei, M., Chai, H., & Jia, F. (2024) · Ocean Engineering · doi:10.1016/j.oceaneng.2024.119274 · Cited by 44
Sensors and Sensor Fusion Methodologies for Indoor Odometry: A Review
Yang, M., Sun, X. S., Jia, F., Rushworth, A. G. A., Dong, X., Zhang, S., Fang, Z., Yang, G., & Liu, B. (2022) · Polymers · doi:10.3390/polym14102019 · Cited by 50
Wang, K., Zhao, G., & Lǚ, J. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2024.3379993 · Cited by 48
“Traditional VSLAM approaches are rooted in geometric models [9], [10], [11], [12], [13], [14], whereas [12], [15] have a specific focus on urban traffic environments for HAV or AVs.”
BackgroundSelf-citationWang, K., Wang, Y., Liu, B., & Chen, J. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3256459 · Cited by 49
“Modern simultaneous location and mapping (SLAM) systems work well in many applications, such as augmented reality and home robotics [99], [100].”
BackgroundSelf-citationSBAS: Salient Bundle Adjustment for Visual SLAM
Wang, K., Ma, S., Ren, F., & Lǚ, J. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3105243 · Cited by 41
Self-citationDeep Learning for Visual SLAM: The State-of-the-Art and Future Trends
Favorskaya, M. N. (2023) · Electronics · doi:10.3390/electronics12092006 · Cited by 43
Agelli, M., Corona, N., Maggio, F., & Moi, P. V. (2024) · Machines · doi:10.3390/machines12110750 · Cited by 37
Aslan, M. F., Durdu, A., & Sabancı, K. (2022) · Measurement · doi:10.1016/j.measurement.2022.111030 · Cited by 36
An Optimization-Based Multi-Sensor Fusion Approach Towards Global Drift-Free Motion Estimation
Wang, K., Cao, C., Ma, S., & Ren, F. (2021) · IEEE Sensors Journal · doi:10.1109/jsen.2021.3064446 · Cited by 29
“Camera is the most commonly used sensor in this field, with many wellknown vision-based algorithms [2], [3].”
Self-citationPalletizing Robot Positioning Bolt Detection Based on Improved YOLO-V3
Zhao, K., Wang, Y., Zuo, Y., & Zhang, C. (2022) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-022-01580-w · Cited by 27
“In recent years, convolutional neural network (CNN) methods have achieved excellent results in many computer vision applications, such as object detection, natural language processing, image recognition, and automatic drive [11–14, 26] and simultaneous localization and mapping [33, 34].”
BackgroundWang, K., Tang, X., Zhao, S., & Zhou, Y. (2021) · Journal of Ambient Intelligence and Humanized Computing · doi:10.1007/s12652-021-02900-y · Cited by 25
Self-citationChamola, V., Chougule, A., Sam, A., Hussain, A., & Yu, F. R. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3362851 · Cited by 26
“In depth estimation domain, a significant amount of research has been done, but there are still many challenges, and issues are open because of its inherent ambiguity [184].”
BackgroundWang, K., Pu, L., Zhang, J., & Lǚ, J. (2022) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2022.3142128 · Cited by 23
“[2], which presents many difficulties for further visual processes.”
BackgroundSelf-citationPVF-DectNet: Multi-modal 3D detection network based on Perspective-Voxel fusion
Wang, K., Zhou, T., Zhang, Z., Chen, T., & Chen, J. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.105951 · Cited by 21
Self-citationWang, K., Pu, L., & Dong, W. (2023) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2023.3308896 · Cited by 21
Self-citationObject-Oriented Semantic SLAM Based on Geometric Constraints of Points and Lines
Ran, T., Yuan, L., Zhang, J., Wu, Z., & He, L. (2022) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2022.3188172 · Cited by 17
“Hence, semantic visual SLAM has been proposed and becomes a hot topic in recent years [3], [4].”
BackgroundXiao, L., Wang, K., Zhou, S. H., & Ma, S. (2021) · Journal of Vibration and Control · doi:10.1177/1077546321999187 · Cited by 17
Self-citationSalient-VPR: Salient Weighted Global Descriptor for Visual Place Recognition
Wang, K., Luo, S., Chen, T., & Lǚ, J. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3219466 · Cited by 15
“This problem is often addressed as visual place recognition (VPR) and is an indispensable module for various applications, ranging from autonomous vehicles [3], [4] and geo-localization [5] to augmented and virtual reality [6].”
BackgroundSelf-citationMDN-VO: Estimating Visual Odometry with Confidence
Kaygusuz, N., Méndez, O., & Bowden, R. (2021) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros51168.2021.9636827 · Cited by 15
“Inspired by this, the field has adapted CNNs to extract and match feature points and to estimate VO [30].”
MethodLiDAR-Based Sensor Fusion SLAM and Localization for Autonomous Driving Vehicles in Complex Scenarios
Dai, K., Sun, B., Wu, G., Zhao, S., Ma, F., Zhang, Y., & Wu, J. (2023) · Journal of Imaging · doi:10.3390/jimaging9020052 · Cited by 14
“With the development of deep learning, related technologies have also been applied to the field of SLAM [32,33].”
MethodMS-VRO: A Multistage Visual-Millimeter Wave Radar Fusion Odometry
Cheng, Y., Jiang, M., & Liu, Y. (2024) · IEEE Transactions on Robotics · doi:10.1109/tro.2024.3400941 · Cited by 14
“Although some works have improved the performance of monocular VO in complex environments [8], [9], its practical application in mobile robots still has limitations to some degree [1].”
BackgroundLi, J., & Luo, J. (2024) · Complex & Intelligent Systems · doi:10.1007/s40747-024-01443-x · Cited by 14
Real-Time Cross-View Image Matching and Camera Pose Determination for Unmanned Aerial Vehicles
Chen, L., Wu, B., Duan, R., & Chen, Z. (2024) · Photogrammetric Engineering & Remote Sensing · doi:10.14358/pers.23-00073r2 · Cited by 14
Towards explainable artificial intelligence in deep vision-based odometry
Ghasemieh, A., & Kashef, R. F. (2024) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2024.109127 · Cited by 13
Where am I heading? A robust approach for orientation estimation of autonomous agricultural robots
Leanza, A., Galati, R., Ugenti, A., Cavallo, E., & Reina, G. (2023) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2023.107888 · Cited by 12
Zhou, Y., Li, X., Li, S., Wang, X., Feng, S., & Tan, Y. (2024) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2024.3400156 · Cited by 17
“Recently, deep learning has significantly propelled the advancement of VSLAM on better accuracy, robustness and dense spatial perception by utilizing higher-level information [2], [3].”
BackgroundLiu, X., He, Y., Li, J., Yan, R., Li, X., & Huang, H. F. (2024) · Sensors · doi:10.3390/s24113388 · Cited by 12
“Although these approaches were successful in location recognition, they heavily depended on viewpoint invariance and robust feature extraction [19].”
BackgroundA Monocular-Visual SLAM System with Semantic and Optical-Flow Fusion for Indoor Dynamic Environments
Chen, W., Shang, G., Hu, K., Zhou, C., Wang, X., Fang, G., & Ji, A. (2022) · Micromachines · doi:10.3390/mi13112006 · Cited by 12
“The calculation and matching of ORB features are fast and have good invariance to viewpoints [22].”
BackgroundÁlvarez-Tuñón, O., Brodskiy, Y., & Kayacan, E. (2023) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2023.3321032 · Cited by 15
“There are also reviews on subsets of SLAM, such as visual odometry [11], [12], loop closure [13], pose graph optimization [14], map building [15], [16], or deep learning, for pose estimation and semantic information [17], [18].”
BackgroundA review of deep learning-based localization, mapping and 3D reconstruction for endoscopy
Zhong, J., Ren, H., Chen, Q. K., & Zhang, H. (2024) · Journal of Micro-Bio Robotics · doi:10.1007/s12213-024-00181-0 · Cited by 12
“Other surveys of visual SLAM such as [9, 34, 37, 40, 47, 88] are mostly targeted for autonomous driving and other open space robotics.”
Cui, X., Sun, Q., Li, Y., Guo, Z. X., & Noureldin, A. M. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3417595 · Cited by 9
“Due to the lower update frequencies of VO and LiDAR, as well as their lack of stability in dynamic environments, they face challenges in estimating heading in scenarios involving rapid movement or complex conditions [15], [16].”
BackgroundWang, Z., Lu, Z., Wei, J., & Qiu, X. (2023) · Applied Sciences · doi:10.3390/app13148443 · Cited by 9
Benchmarking visual SLAM methods in mirror environments
Herbert, P., Wu, J., Ji, Z., & Lai, Y. (2024) · Computational Visual Media · doi:10.1007/s41095-022-0329-x · Cited by 8
“[49–51].”
Wang, K., Zhang, L., Xia, Q., Pu, L., & Chen, J. (2021) · Neural Computing and Applications · doi:10.1007/s00521-021-06609-z · Cited by 8
“Detection is a critical technology in AVs, which can obtain information about the driving scene ahead [40, 42–44].”
BackgroundSelf-citationOptical encoder neural network: a CNN-based optical encoder for robot localization
Patruno, C., Renò, V., Nitti, M., Pernisco, G., & Mosca, N. (2022) · Optical Engineering · doi:10.1117/1.oe.62.4.041402 · Cited by 8
“11,12 A DL-based architecture 13 was proposed to predict velocity and direction changes from stereo images by means of convolutional neural networks (CNNs).”
MethodMollica, G., Legittimo, M., Dionigi, A., Costante, G., & Valigi, P. (2023) · Sensors · doi:10.3390/s23042286 · Cited by 8
Ding, M., Tang, Q., Liu, K., Chen, X., Lu, D., Tian, C., Wang, L., Li, Y., & Wang, G. (2024) · Robotics and Autonomous Systems · doi:10.1016/j.robot.2024.104839 · Cited by 7
Online Learning of Wheel Odometry Correction for Mobile Robots with Attention-based Neural Network
Navone, A., Martini, M., Angarano, S., & Chiaberge, M. (2023) · IEEE International Conference on Automation Science and Engineering (CASE) · doi:10.1109/case56687.2023.10260407 · Cited by 7
A Neural Network Model of the Entorhinal Cortex and Hippocampus for Event-Order Memory Processing
Nakagawa, H., Tateno, K., Takada, K., & Morie, T. (2022) · IEEE Access · doi:10.1109/access.2022.3168715 · Cited by 7
“Furthermore, it is essential to use visual odometry to recognize the depth of the camera image [34], [35].”
MethodAdvanced Visual SLAM and Image Segmentation Techniques for Augmented Reality
Jiang, Y.-R., Tran, T., & Williams, L. (2022) · International Journal of Virtual and Augmented Reality · doi:10.4018/ijvar.307063 · Cited by 4
Surrounding-aware representation prediction in Birds-Eye-View using transformers
Yu, J., Zheng, W., Chen, Y., Zhang, Y., & Huang, R. (2023) · Frontiers in Neuroscience · doi:10.3389/fnins.2023.1219363 · Cited by 6
“This line of research has a wide range of applications in autonomous driving (Ohn-Bar et al., 2020; Yi et al., 2021; Cao et al., 2022a; Wang et al., 2022).”
BackgroundSurvey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization
Brune, M., Meisen, T., & Pomp, A. (2024) · Applied Sciences · doi:10.3390/app14062267 · Cited by 6
Toward Specialized Learning-based Approaches for Visual Odometry: A Comprehensive Survey
Phan, T.-D., & Kim, G.-W. (2025) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-025-02245-0 · Cited by 9
Optimized Deep Learning for LiDAR and Visual Odometry Fusion in Autonomous Driving
Zhang, D., Peng, T., & Loomis, J. S. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3324559 · Cited by 5
“Within this framework, multiview geometry is employed to extract, match, or track feature points to determine motion based on a sequence of images [6].”
MethodMonocular endoscopy images depth estimation with multi-scale residual fusion
Liu, S., Fan, J., Yang, Y., Xiao, D., Ai, D., Song, H., Wang, Y., & Yang, J. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.107850 · Cited by 5
Attention-guided LiDAR segmentation and odometry using image-to-point cloud saliency transfer
Ding, G., İmamoğlu, N., Çağlayan, A., Murakawa, M., & Nakamura, R. (2024) · Multimedia Systems · doi:10.1007/s00530-024-01389-7 · Cited by 5
Cross-View Geo-Localization for Autonomous UAV Using Locally-Aware Transformer-Based Network
Bui, D. V., Kubo, M., & Satō, H. (2023) · IEEE Access · doi:10.1109/access.2023.3317950 · Cited by 5
“The RLV includes popular methods such as Visual Odometry (VO) [4] and Simultaneous Localization and Mapping (SLAM) [5], and many of them have been put into practice [6], [7] [8].”
MethodMSFusion: Multilayer Sensor Fusion-Based Robust Motion Estimation
Li, F., Wei, L., Chen, J., Huang, X. J., & Wang, K. (2023) · IEEE Sensors Letters · doi:10.1109/lsens.2023.3240435 · Cited by 5
“However, just as explored in our previous work [5], [6], [7], [8], there are still some challenges in vision-based SLAM pending to be solved.”
BackgroundSelf-citationCS-SLAM: A Lightweight Semantic SLAM Method for Dynamic Scenarios
Guo, Z., Dong, N., Zhang, Z., Mai, X., & Li, D. (2024) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2024.3462651 · Cited by 4
“With the rapid development of deep learning technology, the prospect of embedded AI methods to integrate robots into everyday life to help with complex everyday tasks is becoming more and more promising [4].”
Zheng, D., Zhao, Y., Tian, W., Zhang, F., Gui, G., & Gui, W. (2025) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2025.3643622 · Cited by 4
“[5], which consists of four residual blocks, with each block’s BN layer guiding the exchange ratio of the channels.”
TQU-SLAM Benchmark Feature-based Dataset for Building Monocular VO
Le, V.-H., Do, H.-S., Phan, V.-N., & Te, T.-H. (2024) · Engineering Technology & Applied Science Research · doi:10.48084/etasr.7611 · Cited by 5
Image Enhancement using GANs for Monocular Visual Odometry
Zubieta, J., Etxeberria-Garcia, M., Zamalloa, M., & Arana-Arexolaleiba, N. (2021) · IEEE International Workshop of Electronics, Control, Measurement, Signals and their Application to Mechatronics (ECMSM) · doi:10.1109/ecmsm51310.2021.9468831 · Cited by 4
“Night driving [4], driving in varying weather conditions [5], underground railways [6], mining [7], or indoor parking [8] are some scenarios where researchers are studying how to take advantage of various types of visual sensors to compute odometry.”
BackgroundA siamese network-based approach for vehicle pose estimation
Zhao, H., Tao, B., Huang, L., & Chen, B. (2022) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2022.948726 · Cited by 4
“DeepVO (Wang et al., 2017; Lee et al., 2021; Wang et al., 2022) is an end-to-end pose estimation network, which uses a deep recurrent convolutional neural network (RCNN) to input a sequence of images and output the corresponding pose directly, without relying on any module in the traditional pose estimation frameworks.”
BackgroundZhao, S., Yan, L., Wang, K., & Chen, J. (2021) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc48978.2021.9564628 · Cited by 3
Self-citationGraph-Laplacian-Processing-Based Multimodal Localization Backend for Robots and Autonomous Systems
Piperigkos, N., Anagnostopoulos, C., Lalos, A. S., Kapsalas, P., & Nguyen, D. V. (2024) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2024.3468712 · Cited by 3
“Apart from traditional geometry-based approaches, deep learning may also used to replace the visual frontend operation for automatically learning effective and robust pixel representations [17].”
SeeTree - a modular, open-source system for tree detection and orchard localization
Brown, J., Grimm, C. M., & Davidson, J. R. (2025) · Smart Agricultural Technology · doi:10.1016/j.atech.2025.101074 · Cited by 3
Federated Meta Learning for Visual Navigation in GPS-denied Urban Airspace
Yüksek, B., Yu, Z., Suri, N., & İnalhan, G. (2023) · IEEE/AIAA Digital Avionics Systems Conference (DASC) · doi:10.1109/dasc58513.2023.10311195 · Cited by 3
“This makes the DL one of the most effective candidate to solve VO problems in challenging conditions [1].”
BackgroundObject-Oriented Semantic SLAM Based on Weighted Scoring of Pose Initialization
Pan, J., & Wang, L. (2023) · China Automation Congress (CAC) · doi:10.1109/cac59555.2023.10450716 · Cited by 3
“Therefore, researchers combine visual SLAM with object detection and proposed semantic visual SLAM [6], [7].”
BackgroundNguyen, T.-H., Le, V.-H., Do, H.-S., Te, T.-H., & Phan, V.-N. (2024) · Future Internet · doi:10.3390/fi16050174 · Cited by 4
“Visual odometry (VO) is applied in many fields, such as [1] autonomous vehicles, unmanned aerial vehicles, underwater robots, space exploration robots, agriculture robots, medical robots, and AR/VR.”
BackgroundMethodInfluentialAtılgan, C., & Mercımek, M. (2025) · IEEE Access · doi:10.1109/access.2025.3571921 · Cited by 2
“Its ability to estimate motion in dynamic environments in real time makes it particularly suitable for various applications in these fields [5].”
Review of deep learning-based visual SLAM: types, approaches, and future work
Zhao, L., Chen, T., Yuan, P., Li, X., & Chen, B. (2025) · Industrial Robot the international journal of robotics research and application · doi:10.1108/ir-04-2025-0137 · Cited by 2
Visual SLAM in the era of heterogeneous intelligence coexistence: A survey
Su, S., He, X., Meng, X., Zhang, Y., Yin, W., Mo, L., & Yu, F. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.132458 · Cited by 2
Wang, Z., & Lu, Z. (2024) · Proceedings of the Institution of Mechanical Engineers Part C Journal of Mechanical Engineering Science · doi:10.1177/09544062231207212 · Cited by 2
Damping Coefficient Optimization for the Articulated System of Virtual Track Trains
Li, C., Ji, Y., Huang, Y., Leng, H., Yang, M., & Ren, L. (2024) · Shock and Vibration · doi:10.1155/2024/8849689 · Cited by 2
Radar-Based Approach for Side-Slip Gradient Estimation
Diener, L., Kalkkuhl, J. C., & Schirle, T. (2024) · SAE technical papers on CD-ROM/SAE technical paper series · doi:10.4271/2024-01-2976 · Cited by 2
Active Illumination for Visual Ego-Motion Estimation in the Dark
Crocetti, F., Dionigi, A., Brilli, R., Costante, G., & Valigi, P. (2025) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra55743.2025.11127536 · Cited by 2
“Deep Learning (DL) approaches leverage Convolutional Neural Networks (CNNs) to mitigate the non-ideal conditions: they can compute features invariant to geometric and photometric changes, including illumination, background, viewpoint, and scale [18].”
A Comparison of Deep Learning-Based Monocular Visual Odometry Algorithms
Jeong, E., Lee, J., & Kim, P. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-19-2635-8_68 · Cited by 2
Deep 4D Automotive Radar-Camera Fusion Odometry with Cross-Modal Transformer Fusion
Lu, S., Zhuo, G., Xiong, L., Zhou, M., & Lu, X. (2023) · SAE International Journal of Advances and Current Practices in Mobility · doi:10.4271/2023-01-7040 · Cited by 2
Semi-Supervised Novelty Detection for Precise Ultra-Wideband Error Signal Prediction
Albertin, U., Navone, A., Martini, M., & Chiaberge, M. (2024) · IEEE International Conference on Automation Science and Engineering (CASE) · doi:10.1109/case59546.2024.10711427 · Cited by 2
“Wheel odometry and visual odometry are standard methods that have been widely investigated in the last decade, although struggling when dealing with long paths and poor repetitive features [7], [8], [9].”
BackgroundHussain, A., & Mehdi, S. R. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4781073 · Cited by 3
AI-Based Visual Odometry Implementation on an Embedded System
Büyüksolak, O., & Güneş, E. O. (2023) · National Conference on Electrical and Electronics Engineering (ELECO) · doi:10.1109/iceee59925.2023.00016 · Cited by 1
“[4] Embedded systems are memory scarce and to implement a neural network on them, memory usage by parameters should be carefully calculated.”
BackgroundMonocular Visual Odometry in Mobile Robot Navigation
Sellak, S., Alj, Y., & Salih-Alj, Y. (2024) · International Symposium on Signal, Image, Video and Communications (ISIVC) · doi:10.1109/isivc61350.2024.10577766 · Cited by 1
“VO plays a crucial role in various computer vision and robotics applications, including Autonomous Mobile Robots, Augmented/ Mixed/ Virtual Reality, and other complex and emerging applications reliant on localization information [2].”
BackgroundAdvanced State Estimation for Multi-Articulated Virtual Track Trains: A Fusion Approach
Lu, Z., Wang, Z., & Luo, X. (2024) · Machines · doi:10.3390/machines12080565 · Cited by 1
Ajewole, O. K., Anjuwon, O., Ezugwu, C. A. K., Adetiba, G., Adewuyi, I. A., & Abiola, J. O. (2024) · IEEE Nigeria International Conference on Disruptive Technologies for Sustainable Development (NIGERCON) · doi:10.1109/nigercon62786.2024.10927138 · Cited by 1
Barai, C., Sarkar, M., Sarkar, U., Mazumder, S., Chandra, A., Samanta, T., & Pandey, H. K. (2026) · Sensors · doi:10.3390/s26072127 · Cited by 1
Mapping the Ocular Surface from Monocular Videos with an Application to Dry Eye Disease Grading
Brahim, I., Lamard, M., Benyoussef, A., Conze, P.-H., Cochener, B., Cornec, D., & Quellec, G. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-16525-2_7 · Cited by 1
“Focusing on monocular visual odometry, the latest and most promising methods are Deep VO [19].”
MethodA Minimal Solution Estimating the Position of Cameras with Unknown Focal Length with IMU Assistance
Yan, K., Yu, Z., Song, C., Zhang, H., & Chen, D. (2024) · Drones · doi:10.3390/drones8090423 · Cited by 1
“Accurately estimating camera relative poses is crucial for various practical applications in drones, such as visual odometry (VO) [1,2], simultaneous localization and mapping (SLAM) [3–5], and structure from motion (SFM) [6,7].”
BackgroundLF2SLAM: Learning-based Features For visual SLAM
Legittimo, M., Crocetti, F., Fravolini, M. L., Mollica, G., & Costante, G. (2024) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros58592.2024.10801935 · Cited by 1
“Specifically, deep learning models that leverage CNNs can compute features invariant to geometric and photometric changes, including illumination, background, viewpoint, and scale [24].”
Semantic-aware dynamic scene tracking and reconstruction for mobile robots leveraging visual SLAM
Liu, T., & Luo, J. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-07343-2 · Cited by 1
“Currently, vSLAM for dynamic environments [5, 6] is mainly categorized into traditional methods and deep learning-based methods.”
Self-Supervised Deep Visual Stereo Odometry with 3D-Geometric Constraints
Zach, J., & Stelldinger, P. (2025) · International Conference on PErvasive Technologies Related to Assistive Environments (PETRA) · doi:10.1145/3733155.3733194 · Cited by 1
“Although classical geometry-based methods have been demonstrated to be highly accurate in favourable circumstances, they frequently prove ineffective in challenging conditions, such as textureless areas or dynamic environments [14].”
Automated Unmapped Forest Path Navigation of Mobile Rover Using Neural Networks
Jovičić, N. (2021) · Robotica & Management · doi:10.24193/rm.2021.2.3 · Cited by 1
Salient Bundle Adjustment for Visual SLAMPreprint
Wang, K., Ma, S., Chen, J., & Lǚ, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.11863
“S LAM means to estimate the pose of the robot in unknown environments, simultaneously construct a model of the environment that the sensors are perceiving [1], [2].”
BackgroundSelf-citation2022 Index IEEE Transactions on Cognitive and Developmental Systems Vol. 14
(2022) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2022.3228651
Bilal, M., Hanif, M. S., Munawar, K., & Al-Saggaf, U. M. (2023) · IEEE Access · doi:10.1109/access.2023.3284463
“On the other hand, deep VO methods provide promising results in challenging environments but perform inferior to their geometry-based counterparts, especially in scenarios favorable to geometry-based methods [2].”
BackgroundMethodInfluentialUnsupervised Full Transformer for Pose, Depth and Optical Flow Joint Learning
Liu, X., & Zhang, T. (2023) · International Conference on Automation, Control and Robotics Engineering (CACRE) · doi:10.1109/cacre58689.2023.10208360
“The learning-based method can be divided into two aspects, supervised and unsupervised, according to whether there is GT value in the training phase [3].”
MethodVisual Odometry Based on Handcrafted and Learned CNN Feature Points
Gao, Y., Zhu, H., Zhang, J., & Li, B. (2023) · China Automation Congress (CAC) · doi:10.1109/cac59555.2023.10450197
“Feature extraction is one of the core issues of feature-based VO, and there are two main approaches in feature extraction: one is based on handcrafted feature design, and the other is based on learning CNN [12].”
MethodHybrid Visual Odometry Algorithm Using a Downward-Facing Monocular Camera
Al‐Hadithi, B. M., Thomas, D., & Pastor, C. (2024) · Applied Sciences · doi:10.3390/app14177732
“However, there are still many challenges to solve in order to achieve a robust and accurate VO using these techniques [24].”
Continuous and generalized visual odometry based on FLANN_PSC-RANSAC_LSTM
Li, D., Feng, R., Wu, F., Zhao, J., & Zhao, J. (2024) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-232279
“…algorithms have been in the process of development and achieved good results, which can be divided into the feature-based method and direct method, both of them can be applied in different image categories, The final results have been proven to provide more accurate and effective results [5].”
Beyond Photometric Constraints: Epipolar-Based Self-Supervised Learning for Visual Odometry
Islam, T., & Wang, Y. (2024) · International Conference on Computers and Artificial Intelligence Technology (CAIT) · doi:10.1109/cait64506.2024.10962878
“Traditional VO methods have largely focused on feature-based and direct approaches [4].”
Towards Low-Cost Real-Time Seafloor Featureless Visual Odometry
Beason, J., Casagrande, D. S., & Roman, C. (2024) · IEEE/OES Autonomous Underwater Vehicles · doi:10.1109/auv61864.2024.11030794
“As a result, recent work increasingly adopts deep learning as an alternative or a supplement to geometric methods [12].”
P‐3.6: A Survey of Visual Odometry Based on Deep Learning
zhang, P., & Li, W. (2025) · SID Symposium Digest of Technical Papers · doi:10.1002/sdtp.18956
Metadata
- Authors
- Wang, K., Ma, S., Chen, J., Ren, F., & Lu, J.
- Affiliations
- Chongqing University (CN)
- Chongqing Normal University (CN)
- Changan Auto Company, Intelligent Vehicle R&D Institute, Chongqing, China
- Ford Motor Company (United States) (US)
- Venue
- IEEE Transactions on Cognitive and Developmental Systems
- Published
- 2022-03
- Volume
- 14
- Issue
- 1
- Pages
- 35-49
- ISSN
- 2379-8920 (ISSN-L)
- 2379-8939
- OpenAlex
- W3083208433
- Cited by
- 114
- References
- 127
- Citations by year
- 2026: 10 · 2025: 15 · 2024: 35 · 2023: 25 · 2022: 19
- FWCI
- 8.11
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Cognitive and Developmental Systems (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Natural Science Foundation of China (51605054)
- National Key Research and Development Program of China (SQ2020YFF0410766)
- Natural Science Foundation of Chongqing (cstc2020jcyj-msxmX0575)
- Chongqing Technology Innovation and Application Development Project (cstc2020jscxmsxmX0109)
- State Key Laboratory of Vehicle NVH and Safety Technology (NVHSKL-202008, NVHSKL-202010)
- Fundamental Research Funds for the Central Universities (2020CDJ-LHZZ-042)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Robotics and Sensor-Based Localization
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Physical Sciences › Earth and Planetary Sciences › Geology › 3D Surveying and Cultural Heritage
- Keywords
- visual odometry, deep learning, relative localization, depth estimation, feature matching, pose estimation, mobile robots, medical robots, augmented reality, virtual reality
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
