The DOI matches its Crossref record; 29 of the 29 references it cites were found in a registry or an index.
DeepVO: Towards end-to-end visual odometry with deep Recurrent Convolutional Neural Networks
Open accessAbstract
This paper studies monocular visual odometry (VO) problem. Most of existing VO algorithms are developed under a standard pipeline including feature extraction, feature matching, motion estimation, local optimisation, etc. Although some of them have demonstrated superior performance, they usually need to be carefully designed and specifically fine-tuned to work well in different environments. Some prior knowledge is also required to recover an absolute scale for monocular VO. This paper presents a novel end-to-end framework for monocular VO by using deep Recurrent Convolutional Neural Networks (RCNNs). Since it is trained and deployed in an end-to-end manner, it infers poses directly from a sequence of raw RGB images (videos) without adopting any module in the conventional VO pipeline. Based on the RCNNs, it not only automatically learns effective feature representation for the VO problem through Convolutional Neural Networks, but also implicitly models sequential dynamics and relations using deep Recurrent Neural Networks. Extensive experiments on the KITTI VO dataset show competitive performance to state-of-the-art methods, verifying that the end-to-end Deep Learning technique can be a viable complement to the traditional VO systems.
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.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“More surprisingly, they show that the welltrained networks enable scale-consistent predictions over long videos, while the accuracy is still inferior to traditional methods because of ignoring geometric information.”
From DF-VO: What Should Be Learnt for Visual Odometry? (2021) · cites this paper as Wang et al. 2017
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“On the opposing side, end-to-end approaches can estimate the 6 degrees of freedom (DoF) poses directly from a sequence of RGB images, that is, it does not need engineering effort to design the modules in the traditional VO pipeline [4].”
From Transformer-Based Model for Monocular Visual Odometry: A Video Understanding Approach (2023) · cites this paper as [4]
The cited source says what the claim attributes to it.
In the citing paper
“This problem is called monocular visual odometry and often relies on geometric approaches that require considerable engineering effort for a specific scenario.”
From Transformer-Based Model for Monocular Visual Odometry: A Video Understanding Approach (2023) · cites this paper as [4]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“the amount of crater detections that ODN-detector produces is robust to disturbances or modifications of the image quality due to Gaussian noise, shot noise, brightness levels, as well as different illumination conditions.”
From Optical navigation for Lunar landing based on Convolutional Neural Network crater detector (2022) · cites this paper as [16]
Experimental — accuracy not yet measured
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
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 1×
“There have been some popular and powerful DNN architectures, such as VGGNet [22] and GoogLeNet [23], developed for computer vision tasks, producing remarkable performance.”
MethodGoing deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
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 2×
“In this section, we evaluate the proposed end-to-end monocular VO approach on the well-known KITTI VO/SLAM benchmark [3].”
BackgroundMethod(2016) · Cited by 11,679
ORB-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 1×
“Examples include keyframe base PTAM [9] and ORB-SLAM [10].”
BackgroundParallel Tracking and Mapping for Small AR Workspaces
Klein, G., & Murray, D. (2007) · 2007 6th IEEE and ACM International Symposium on Mixed and Augmented Reality · doi:10.1109/ismar.2007.4538852 · Cited by 4,346 · cited 1×
“Examples include keyframe base PTAM [9] and ORB-SLAM [10].”
BackgroundFlownet: Learning optical flow with convolutional networks
Dosovitskiy, A., Fischer, P., Ilg, E., Hausser, P., Hazirbas, C., Golkov, V., Smagt, P. V. D., Cremers, D., & Brox, T. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.316 · Cited by 4,313 · cited 2×
“In order to reduce both the training time and data required to converge, the CNN is based on a pre-trained FlowNet model [24].”
MethodMonoSLAM: Real-time single camera SLAM
Davison, A. J., Reid, I. D., Molton, N. D., & Stasse, O. (2007) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2007.1049 · Cited by 3,959 · cited 1×
“To mitigate this problem, visual Simultaneous Localisation and Mapping (SLAM) or Structure from Motion (SfM) can be adopted to maintain a feature map for drift correction along with pose estimation [8].”
MethodORB-SLAM: A Versatile and Accurate Monocular SLAM System
Mur-Artal, R., Montiel, J. M. M., & Tardos, J. D. (2015) · Zaguan (Universidad de Zaragoza) · Cited by 3,153
Multiple view geometry in computer vision
Opower, H. (2002) · Optics and Lasers in Engineering · doi:10.1016/s0143-8166(01)00145-2 · Cited by 3,044
Flagged first, then by citations
Research works printed but not linked: 1
- zaremba, 2014, Learning to execute
Cited by
Sorted by citations
Items 701–800 of 969
A New Approach to ORB Acceleration Using a Modern Low-Power Microcontroller
Aráez, J., Real, S., & Araújo, Á. (2025) · Sensors · doi:10.3390/s25123796 · Cited by 1
“In contrast, odometry and dead reckoning rely on differential data of the device’s position and orientation, usually gathered with cameras (visual odometry, e.g., [7]) in combination with inertial measurements (e.g., [8]).”
A Survey of Visual SLAM Based on RGB-D Images Using Deep Learning and Comparative Study for VOE
Le, V.-H., & Nguyen, T.-H.-P. (2025) · Algorithms · doi:10.3390/a18070394 · Cited by 1
RadarTrack: Enhancing Ego-Vehicle Speed Estimation with Single-chip mmWave Radar
Sen, A., Chakraborty, S., Tripathy, S., & Chakraborty, S. (2025) · IEEE International Conference on Smart Computing · doi:10.1109/smartcomp65954.2025.00056 · Cited by 1
“These models have been employed across various sensors, including RGB cameras [18], depth cameras [19], [20], IMUs [21], [22], and LiDARs [10].”
Occlusion-Aware Monocular Visual Odometry for Robust Trajectory Tracking
Kang, W., Gai, S., Da, F., & Cai, Z. (2025) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2025.3598609 · Cited by 1
UR-MVO: Robust Monocular Visual Odometry for Underwater Scenarios
Barhoum, Z. A., Maalla, Y., Daher, S., Topolnitskii, A., Mahmoud, J., & Kolyubin, S. A. (2025) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra55743.2025.11128500 · Cited by 1
“DeepVO [25] combines CNN for feature extraction and LSTM for modeling temporal dependencies between features to estimate pose increments.”
DVLO4D: Deep Visual-Lidar Odometry with Sparse Spatial-Temporal Fusion
Liu, M., Yang, M. Y., Liu, J., Zhang, Y., Li, J., Elberink, S. O., George Vosselman, A., & Cheng, H. (2025) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra55743.2025.11127668 · Cited by 1
“DeepVO [31] employs deep recurrent neural networks to model temporal dependencies in sequential data.”
Simsek, E., & Özyer, B. (2025) · Image and Vision Computing · doi:10.1016/j.imavis.2025.105732 · Cited by 1
VSTFusion-VO: Monocular Visual Odometry with Video Swin Transformer Multimodal Fusion
Hsu, C. T., Lin, H.-C., Jhong, S.-Y., Hsu, H.-C., Hong, M.-X., & Chen, Y.-Y. (2025) · International Conference on Advanced Robotics and Intelligent Systems (ARIS) · doi:10.1109/aris66143.2025.11163438 · Cited by 1
“To overcome these limitations, learning-based VO approaches [4], [5] directly estimate poses from raw RGB inputs, reducing dependence on manual design and improving adaptability.”
Tang, J., Yu, J., Xiang, Y., Min, X., Xu, Y., Dong, Y., & Wang, Y. (2025) · IEEE International Conference on Automation Science and Engineering (CASE) · doi:10.1109/case58245.2025.11164141 · Cited by 1
“These methods encompass various techniques, such as direct methods that directly estimate the camera motion using pixel intensities [4]–[6], and feature-based approaches that rely on detecting and tracking distinct points, lines or planes in the environment [7]–[9].”
Zhou, K., Yu, Z., Zhou, X., Tan, P., Yin, Y., & Luo, H. (2025) · Electronics · doi:10.3390/electronics14204126 · Cited by 1
TSFNet: A Two-Stage Fusion Network for Visual–Inertial Odometry
Wang, S., Liang, Y., Lin, J., Gan, Y., Zhong, M., Yin, X., & Bao, P. (2025) · Mathematics · doi:10.3390/math13233842 · Cited by 1
Self-citationPirozzo, B. M., Paula, M. D., Villar, S. A., de Benito, C., Acosta, G. G., & Picos, R. (2025) · IET Cyber-Systems and Robotics · doi:10.1049/csy2.70028 · Cited by 1
Real-Time FPGA-Based Hardware-Algorithm Co-Design for Monocular Visual Odometry on Edge Devices
Huang, L.-Y., Hsieh, Y.-K., & Chien, S. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3664685 · Cited by 1
“More recent approaches, including DeepVO [16], employ Long Short-Term Memory (LSTM) networks to capture sequential dependencies inherent in pose estimation.”
Principled Uncertainty Propagation for Stereo Visual Odometry
Buonomo, F. V., Andolfo, S., & Genova, A. (2026) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-026-02358-0 · Cited by 1
“While learning-based methods can excel when large curated datasets exist [34, 46, 47], their dependence on data distribution and domain alignment can hinder deployment in data-scarce settings typical of planetary exploration.”
Huang, X., Zhu, Z., Zhang, X., Cao, W., & Yu, S. (2025) · International Conference on Electronic Information Engineering and Computer Communication (EIECC) · doi:10.1109/eiecc67963.2025.11409644 · Cited by 1
“…blur, or depth discontinuities, and they often require careful parameter tuning across environments.Deep learning based VO models such as DeepVO [13] further improve robustness by learning a motion prior from data, but most of them are designed for monocular RGB input and are trained on outdoor…”
One Step Closer: Creating the Future to Boost Monocular Semantic Scene Completion
Lu, H., Su, Y., Zhang, X., & Hu, H. (2025) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc60802.2025.11423618 · Cited by 1
“The complete loss function is formulated as: For FuturePoseNet, following [38], we supervise it using MSE loss.”
Implicit Point-to-Voxel LiDAR-IMU SLAM
Dong, Y., Xu, E., Yang, J., Qiu, S., Wang, J., & Han, B. (2026) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2026.3683446 · Cited by 1
“It should be emphasized that the proposed method is not an end-to-end pose learning framework [19], [20], [21], [22], [23] that directly regresses robot motion from sensor inputs.”
Robust Monocular Visual Odometry With Temporal Augmentation and Dynamic-Aware Attention
Lin, H.-C., Jhong, S.-Y., Lin, Y.-H., Chang, C.-A., Fang, J.-K., & Chen, Y.-Y. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3686397 · Cited by 1
“To ensure a fair comparison with state-of-the-art learning-based methods (e.g., DeepVO [14] and SWformer-VO [23]), we strictly follow the standard evaluation split established in the literature.”
InfluentialDeep Learning and Geometry-based Image Localization Enhanced by Bluetooth Signals
Ishihara, T., Kitani, K., Asakawa, C., & Hirose, M. (2018) · Journal of Information Processing · doi:10.2197/ipsjjip.26.707 · Cited by 1
Toward Accurate Motion Estimation with Affine Correspondences
Alyousefi, K. (2020) · Digital Collections of Colorado (Colorado State University) · Cited by 1
Localising In Complex Scenes Using Balanced Adversarial Adaptation
Avraham, G., Zuo, Y., & Drummond, T. W. (2020) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv50981.2020.00116 · Cited by 1
HighStitch: High Altitude Georeferenced Aerial Images Stitching for Rocking Telephoto Lens
Zhao, Y., Xu, S., Bu, S., Jiang, H., & Han, P. (2021) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2021.3124745 · Cited by 1
MOAC: Multi-level Perception Optimizer Based on Dual Augmented Cost for Structure- from-Motion
Wu, P., Li, G., & Li, T. H. (2022) · IEEE International Conference on Multimedia Information Processing and Retrieval (MIPR) · doi:10.1109/mipr54900.2022.00031 · Cited by 1
“Early methods used neural networks to regress results directly from the input images [19, 20, 26, 22], which did not work well due to the lack of geometric conFigure 1.”
MethodSpatial Consciousness Model of Intrinsic Reward in Partially Observable Environments
Ni, Z., Jin, Y., Liu, P., & Zhao, W. (2022) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-022-01771-5 · Cited by 1
“The navigation modes of vision-based SLAM[18, 19] and filtering methods[20, 21] can effectively cope with tasks requiring long-term work in a fixed environment.”
Distributed Estimation of Sensor Networks Based on LSTM-Kalman Filter
Kuang, S.-P., & Zhang, Y. (2022) · International Conference on Control, Automation, Robotics and Vision (ICARCV) · doi:10.1109/icarcv57592.2022.10004275 · Cited by 1
“In recent years, there has been an explosion of the application of deep neural networks for motion estimation [3]– [5].”
BackgroundSilvestrini, S., Cassinis, L. P., Hinz, R., González-Arjona, D., Tipaldi, M., Visconti, P., Corradino, F., Pesce, V., & Colagrossi, A. (2023) · Elsevier eBooks · doi:10.1016/b978-0-323-90916-7.00015-9 · Cited by 1
DAVO:A Monocular Visual Odometry Method Based on Dual Attention
Li, J., & Zheng, B. (2022) · International Conference on Algorithms, Computing and Artificial Intelligence (ACAI) · doi:10.1145/3579654.3579704 · Cited by 1
“DeepVO [20] combines convolutional neural networks (ConvNet) and recurrent neural networks (RNN) for end-to-end learning of visual odometry.”
BackgroundRamasamy, K., Chandrasekar, M., & Thamilkolunthu, S. (2023) · High Temperature Materials and Processes · doi:10.1515/htmp-2022-0273 · Cited by 1
Liang, J. T. Y., & Kanji, T. (2023) · International Conference on Pattern Recognition Systems (ICPRS) · doi:10.1109/icprs58416.2023.10179060 · Cited by 1
“This problem has been extensively studied in computer vision, with various approaches such as obstacle detection [1], free space detection [2], and 3D reconstruction [3, 4, 5].”
BackgroundSelf-supervised Visual Odometry Based on Geometric Consistency
Song, R., Liu, J., Liao, K., Xiao, Z., & Yan, B. (2023) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas46773.2023.10181492 · Cited by 1
“The supervised monocular VOs built on DL networks such as CNNs [14] and RNNs [9] [10] have achieved promising performance.”
BackgroundOverview of Visual SLAM Technology: From Traditional to Deep Learning Methods
Huang, K. (2023) · Advances in Computer Signals and Systems · doi:10.23977/acss.2023.071011 · Cited by 1
Lesiv, E. A., Vytovtov, K. A., & Barabanova, E. A. (2023) · International Conference on Information, Control, and Communication Technologies (ICCT) · doi:10.1109/icct58878.2023.10347079 · Cited by 1
“The solution of the visual odometry problem based on neural network algorithms is proposed in [8, 9].”
MethodUniversal Adversarial Attacks for Visual Odometry Systems
Xie, X., Liao, L., Yu, Y., Guo, D., & Liu, H. (2023) · IEEE International Conference on Development and Learning (ICDL) · doi:10.1109/icdl55364.2023.10364418 · Cited by 1
“Deep neural networks have been employed for tasks such as feature extraction and motion estimation, exemplified by DeepVO [1], Mon-oDepth2 [2], Depth VO Feat [3], and CNN-SVO [4], These methods have demonstrated significant improvements in both Xijin Xie, Longlong Liao and Yuanlong Yu are with…”
MethodA Distributed Deep Learning System With Controlled Intermediate Representation
Xiao, Y., Wang, Y., Huang, Z., Shen, F., & Qin, F. (2023) · IEEE Smart World Congress (SWC) · doi:10.1109/swc57546.2023.10448716 · Cited by 1
“…deep learning has made rapid development in various fields, such as Computer Vision [1] [2], Natural Language Processing [3] [4], Visual Odometry [5] [6], etc. Benefited by the efficient data feature extraction and analysis capability of Deep Neural Network [7](DNN), more advanced sensors with…”
MethodA review of Multisensor Fusion techniques used in the development of Mobile Robotic Systems
Chandrasekaran, B. (2023) · International Conference on Robotics and Automation Engineering (ICRAE) · doi:10.1109/icrae59816.2023.10458543 · Cited by 1
Tabrizi, H. B., & Crick, C. (2023) · International Conference on Computational Science and Computational Intelligence (CSCI) · doi:10.1109/csci62032.2023.00235 · Cited by 1
“The underlying features of the task have been learned using deep learning methods to achieve better results than traditional approaches.”
Gurumurthy, S., Ram, K., Chen, B., Manchester, Z. R., & Kolter, J. Z. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.02597 · Cited by 1
“These tasks have a long history of being tackled using geometry-based opti-mizaton [23, 24, 35, 41], and in the last decade, using deep networks to directly map inputs to poses [42, 56, 58, 61].”
BackgroundOnline lane mapping based on multi-sensor SLAM and Catmull–Rom splines
Gao, T., Zou, D., Chen, C. P., Wu, X., & Hu, H. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/ada8c8 · Cited by 1
“DeepVO [18] can compute the pose directly from a sequence of raw images without using any modules from traditional SLAM frameworks.”
Chen, S., Lu, L., Chen, D., Chen, D., & Kong, D. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/adb771 · Cited by 1
“Unlike [5] and [6], which use RGB as input, we use optical flow.”
InfluentialGG-SLAM: An Object-Level SLAM System Using Gaussian Grouping and Splatting
Jung, H., & Kim, G.-W. (2025) · The Journal of Korea Robotics Society · doi:10.7746/jkros.2025.20.1.084 · Cited by 1
Beyond Handcrafted Features: A Deep Learning Framework for Optical Flow and SLAM
Kazi, K., Khizer, A. N., Memon, F., Memon, A. R., & Iqbal, M. (2025) · Journal of Imaging · doi:10.3390/jimaging11050155 · Cited by 1
MPVO: Motion-Prior Based Visual Odometry for PointGoal Navigation
Paul, S., Roychoudhury, R. D., & Bhowmick, B. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-91569-7_9 · Cited by 1
RAEM-SLAM: A Robust Adaptive End-to-End Monocular SLAM Framework for AUVs in Underwater Environments
Wu, Y., Li, Y., Luo, W., & Ding, X. (2025) · Drones · doi:10.3390/drones9080579 · Cited by 1
Analysis of the Effect of Sensors for End-to-End Machine Learning Odometry
Rodriguez-Peral, C. M., & Peña, D. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-11024-6_6 · Cited by 1
“DeepVO and VISO2 M results are taken from [16].”
BackgroundMethodAdaptive Multi-Scale Tracking Target Algorithm through Drone
He, Q., Shao, X., Chen, W., Li, X., Xiao, Y. A. N. G., & Sun, T. (2019) · IEICE Transactions on Communications · doi:10.1587/transcom.2018drp0040 · Cited by 1
“Tracking algorithm includes from early Mean Shift al-gorithm [9]–[11], particle filter algorithm [12]–[14], support vector machine (SVM) [15]–[17] to multiple instance learning algorithm [18]–[20].”
MethodAPAC-Net: Unsupervised Learning of Depth and Ego-Motion from Monocular Video
Lin, R., Lu, Y., & Lu, G. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-36189-1_28 · Cited by 1
Monocular Visual Odometry Using Unsupervised Deep Learning
Liu, F., Liu, Z., & Wu, Q. (2019) · Chinese Automation Congress · doi:10.1109/cac48633.2019.8996257 · Cited by 1
“Our monocular VO system uses LSTM (Long Short-Term Memory) like Wang and Liang [5][6] instead of standard RNN, since LSTM solves long-term dependencies existing in the latter.”
BackgroundMethodVision-Based Pose Optimization Using Learned Metrics
Li, S., Xian, Y., Zhang, T., Li, B., Zhang, D., & Guo, W. (2020) · IEEE Access · doi:10.1109/access.2020.3021824 · Cited by 1
“The learning methods are divided into supervised learning with pose labels [30], [31] and unsupervised learning without pose labels [32]–[34].”
BackgroundSelf-Supervised Visual Odometry with Ego-Motion Sampling
Slinko, I., Vorontsova, A., Zhukov, D., Barinova, O., & Konushin, A. (2020) · International Conference on Video, Signal and Image Processing (VSIP) · doi:10.1145/3442705.3442706 · Cited by 1
“Deep learning-based methods of visual odometry emerged in recent years [1, 10, 17, 34, 35, 37, 42] and now they achieve state-of-the-art results [10] on the KITTI [11] benchmark among all monocular-view methods.”
BackgroundMethodInfluentialLearning Visual Measurement Model for Kalman Filter with Deep Convolutional Neural Networks
Chen, Z., Ji, L., Li, Z., & de Silva, C. W. (2020) · International Conference on Digital Home (ICDH) · doi:10.1109/icdh51081.2020.00065 · Cited by 1
A Consistency-Based Loss for Deep Odometry Through Uncertainty Propagation
Damirchi, H., Khorrambakht, R., Taghirad, H. D., & Moshiri, B. (2023) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48891.2023.10160954 · Cited by 1
“To compare against deep learning based approaches, we chose UA-VO [10], ESP-VO [7], DeepVO [1] 1The code will be available upon acceptance 11759 Authorized licensed use limited to the terms of the applicable license agreement with IEEE.”
BackgroundMethodInfluentialMarine Vehicles Localization Using Grid Cells for Path Integration
Carlucho, I., Bailey, M. F., Paula, M. D., & Barbălată, C. (2021) · MTS/IEEE OCEANS Conference (OCEANS) · doi:10.23919/oceans44145.2021.9706050 · Cited by 1
Unsupervised Learning Of Visual Odometry Using Direct Motion Modeling
Andrei, S. S., & Gady Agam, B. (2021) · Proceedings - International Conference on Image Processing · doi:10.1109/icip42928.2021.9506572 · Cited by 1
“Learning-based approaches can be divided in two main categories: Supervised methods [6, 7, 8, 9] which use ground-truth pose and/or depth as a supervisory signal; and unsupervised methods [10, 11, 12, 13, 14, 15].”
MethodAttention Mechanism-based Monocular Depth Estimation and Visual Odometry
Zhang, Q., Lin, D., Ren, Z., Kang, Y., Wu, F., & Cheng, J. S. (2021) · IEEE International Conference on Real-time Computing and Robotics (RCAR) · doi:10.1109/rcar52367.2021.9517422 · Cited by 1
Neural-based Visual Odometry Trained in a Virtual Enviroment for a Mobile Robot Navigation
Diane, S. A. K., & Lesiv, E. A. (2021) · International Conference Management of Large-Scale System Development (MLSD) · doi:10.1109/mlsd52249.2021.9600249 · Cited by 1
Encoding A Mathematically Faithful DeepVIO Solution
Khan, A. A. N., & Aouf, N. (2021) · International Conference on Control, Automation and Systems (ICCAS) · doi:10.23919/iccas52745.2021.9649964 · Cited by 1
Wang, Y., & Peng, X. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-16-9423-3_39 · Cited by 1
Wang, Y., Fu, Y., Zheng, R., Wang, L., & Qi, J. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-16-9423-3_38 · Cited by 2
Higher accuracy self-supervised visual odometry with reliable projection
Zhou, S. S., Yang, Z., Zhu, M., Li, H., Serikawa, S., Mizumachi, M., & Zhang, L. (2022) · Artificial Life and Robotics · doi:10.1007/s10015-022-00766-7 · Cited by 1
“Later, Deepvo [23] proposed a recurrent ConvNet for learning pose from video sequences.”
BackgroundSemantic Tagging of CAN and Dash Camera Data from Naturalistic Drives
Sanborn, K., Richardson, A., & Sprinkle, J. M. (2022) · ACM/IEEE International Conference on Cyber-Physical Systems (ICCPS) · doi:10.1109/iccps54341.2022.00047 · Cited by 1
An End-To-End Visual Odometry based on Self-Attention Mechanism
Cao, R., Wang, Y., Yan, K., Chen, B., Ding, T., & Zhang, T. (2022) · IEEE International Conference on Power, Intelligent Computing and Systems (ICPICS) · doi:10.1109/icpics55264.2022.9873538 · Cited by 1
“In this subsection, we show the comparison of the invented algorithm in the paper and its comparative algorithms DeepVO, SfMLearner and VISO2-M on the Sequence 04, 05, 06, 07, 10.”
MethodInfluentialLyu, Y., & Zhang, C. G. (2022) · Computer Science and Information Systems · doi:10.2298/csis220320035l · Cited by 1
SCVO: Scale-Consistent Depth and Pose for Unsupervised Visual Odometry
Zhang, H., Yang, Y., Li, J., & Li, W. (2022) · Chinese Control Conference (CCC) · doi:10.23919/ccc55666.2022.9902340 · Cited by 1
“At present, the mainstream supervised VO methods include DeepVO [12], ESP-VO [13], VIO-NET [14], DA-VO [15],etc. Supervised VO learns from large amounts of labeled sensor data, which means huge human effort on labeling data.”
MethodRobust Vision SLAM Based on YOLOX for Dynamic Environments
Yan, H., Zhou, X., Liu, J., Yin, Z., & Yang, Z. C. (2022) · IEEE International Conference on Communication Technology (ICCT) · doi:10.1109/icct56141.2022.10073383 · Cited by 1
Lightweight Deep Learning Algorithm for Visual Odometry
Zhang, D., Peng, T., Liu, R., Zhao, D., Sun, K., & Loomis, J. S. (2023) · IEEE National Aerospace and Electronics Conference (NAECON) · doi:10.1109/naecon58068.2023.10365901 · Cited by 1
“DeepVO [9] is a deep learning-based visual odometry algorithm that learns to estimate the camera pose from a single monocular image sequence.”
BackgroundTraining Deep SLAM on Single FramesPreprint
Slinko, I., Vorontsova, A., Zhukov, D., Barinova, O., & Konushin, A. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1912.05405 · Cited by 1
“Deep learning-based methods for visual odometry and SLAM [32, 33, 41, 14, 35] have evolved over the last years and are able to compete with classical geometry-based methods on the datasets with predominant motions along plain surfaces like KITTI[12] and Malaga[3].”
BackgroundMethodHausberg, J., Ishikawa, R., Roxas, M., & Oishi, T. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.07008 · Cited by 1
Vision based navigation for extra terrestrial lava tubes exploration
Debeunne, C. (2024) · doi:10.70675/3285aa94z09f5z4b73za1a6zcd22eaaf50a6 · Cited by 1
Crowdsourced 3D Mapping: A Combined Multi-View Geometry and Self-Supervised Learning Approach
Chawla, H., Jukola, M., Brouns, T., Arani, E., & Zonooz, B. (2020) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros45743.2020.9341243 · Cited by 6
“Supervised learning approaches predict monocular depth [32]– [34] and ego-motion [10], [35] using ground truth depths and trajectories, respectively.”
BackgroundDeep Visual Odometry Methods for Mobile RobotsPreprint
Shabbir, J., & Kruezer, T. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1807.11745
“For this reason, the process is quite a challenging task that is related to mapping and simultaneous localization whose main aim is to generate the road map from a stream of visual data [3].”
BackgroundOn Exploiting Per-Pixel Motion Conflicts to Extract Secondary Motions
Babu, B. P. W., Yan, Z., Ye, M., & Ren, L. (2018) · Proceedings - International Symposium on Mixed and Augmented Reality, ISMAR · doi:10.1109/ismar.2018.00028
“However, after over 30 years of research, the robustness of SLAM still needs improvement for its application to real-life conditions [8].”
BackgroundInfluentialUse of First and Third Person Views for Deep Intersection ClassificationPreprint
Takeda, K., & Kanji, T. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1901.07446
“The distance from the intersection is greater in TPV than in FPV, and the distance to the intersection (D2I) can be measured by VO [8] or egomotion regression [1].”
Long-Term Vehicle Localization by Recursive Knowledge DistillationPreprint
Tomoe, H., & Kanji, T. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.03551
“…at maximum possible vehicle exploration of the entire environment, while keeping track of the vehicle’s global position (e.g., using pose tracking [6] and relocation [7]), to collect mapped images that have global viewpoint information, and optionally, the collected data may be further…”
LIDAR Odometry with Joint Geometric and Appearance Landmarks
Skikos, B. A. P. (2019) · UWSpace (University of Waterloo)
Improving drone localisation around wind turbines using monocular model-based trackingPreprint
Moolan-Feroze, O., Karachalios, K., Nikolaidis, D., & Calway, A. D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.10474
“This method is also used in [10], where the authors state the recurrent layers are able to learn the motion dynamics of the system over a period of time.”
Benchmark for Deep Learning based Visual Odometry and Monocular Depth Estimation
Choi, H. (2019) · The Journal of Korea Robotics Society · doi:10.7746/jkros.2019.14.2.114
Camera Pose Estimation Method Based on Deep Neural Network
Qing, T. X., Wu, F., & Tao, Z. Y. (2019) · International Conference on Deep Learning Technologies (ICDLT) · doi:10.1145/3342999.3343014
“DeepVO [12] uses time-series RGB images for pose estimation with RCNN network structure.”
MethodEnd-to-end Network for Monocular Visual Odometry Based on Image Sequence
Yao, M., & Quan, H. (2019) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2019.8852165
Li, Q., Wang, C., Chen, S., Li, X. S., Wen, C., Cheng, M., & Li, J. (2019) · The international archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprs-archives-xlii-2-w13-1681-2019
“Due to the difference in scale and units between the translational and rotational pose components, previous works (Kendall et al., 2015, Wang et al., 2017) gave a weight regularizer λ to the rotational loss to jointly learn the 6-DoF pose.”
MethodCombined Learned and Classical Methods for Real-Time Visual Perception in Autonomous Driving
Aladem, M. (2020) · Deep Blue (University of Michigan)
Spatio-Temporal and Geometry Constrained Network for Automobile Visual Odometry
Liu, H., Peng, W., Huang, W., Hua, G., & Meng, F. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9052903
“On average, our method is superior to other monocular learning or geometry based methods in both translation and rotation, with the errors decreased by 30% and 67% compared with DeepVO [20].”
ResultEstimation of Absolute Scale in Monocular SLAM Using Synthetic Data
Rukhovich, D. D. (2020) · PROGRAMMNAYA INGENERIA · doi:10.17587/prin.11.86-95
Voice Control of Fetch Robot Using Amazon Alexa
Liu, P. (2020) · VTechWorks (Virginia Tech)
Ego-Motion Estimation Using Affine Correspondences
Alyousefi, K., & Ventura, J. (2020) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-43020-7_70
Semantics-Driven Unsupervised Learning for Monocular Depth and Ego-Motion EstimationPreprint
Wei, X., Feng, J., & Zhou, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.04371
“[46] proposed an end-to-end monocular framework, DeepVO, which not only automatically learns effective feature representation for the VO problem through CNN, but also implicitly models sequential dynamics and relations using deep RNN.”
BackgroundQiu, J. (2020) · The Sydney eScholarship Repository (The University of Sydney)
Kweon, I. S., & Rameau, F. (2020) · Springer eBooks · doi:10.1007/978-3-030-03243-2_848-1
Feasibility Study on Optical Image Modulation Based Parallax Generator for Monocular Visual Odometry
Lee, M., Cha, M., Park, C. S., & Kim, K.-S. (2020) · International Conference on Control, Automation and Systems (ICCAS) · doi:10.23919/iccas50221.2020.9268376
“Recently, the results of end-to-end learning using the Deep Recurrent Convolutional Neural Network, which does not depend on such depth information, have been disclosed [3].”
BackgroundJoint Estimation of Depth and Pose with IMU-assisted Photometric Loss
Yang, X., Feng, Y., Li, J., Chen, Z., Sun, Z., & Nie, X. (2020) · IEEE International Conference on Unmanned Systems (ICUS) · doi:10.1109/icus50048.2020.9274869
Salient Bundle Adjustment for Visual SLAMPreprint
Wang, K., Ma, S., Chen, J., & Lǚ, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.11863
“Deep VO [58] use RNN to model the relationship of motion dynamic and image sequences.”
MethodSpacecraft Relative-Kinematics State Estimation using Conditional Normalizing Flows
Heintz, A. M., Peck, M. A., & Pellumbi, R. (2021) · AIAA SciTech Forum · doi:10.2514/6.2021-1954
Deep Weakly Supervised PositioningPreprint
Wang, R., Xu, X., Ding, L., Huang, Y., & Feng, C. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.04866
“DeepVO [35] and VidLoc [36] propose to use a combined CNN and LSTM networks for monocular VO by taking the advantage that CNN and LSTM are capable to extract features from”
MethodThe Realtime Indoor Localization Unmanned Aerial Vehicle
Zhou, Y., Yu, Z., & Ma, Z. (2021) · Communications in computer and information science · doi:10.1007/978-981-16-2336-3_5
LIFT-SLAM: a deep-learning feature-based monocular visual SLAM method
Bruno, H. M. S., & Colombini, E. L. (2020) · doi:10.5753/wtdr_ctdr.2020.14954
RAM-VO: Less is more in Visual OdometryPreprint
Cleveston, I., & Colombini, E. L. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.02974 · Cited by 1
“In 2017, Wang et al. [13] proposed DeepVO, which is an end-to-end monocular architecture capable of extracting visual features directly from the input raw images with a CNN; and determining temporal relation with an LSTM.”
Algorithm Design of Semantic Map Construction for Dynamic Scene
Zheng, F., Wang, X., Xia, F., & Shi, Z. (2021) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-030-81007-8_17
Incorporating Learnt Local and Global Embeddings into Monocular Visual SLAMPreprint
Huang, H., Ye, H., Sun, Y., Wang, L., & Liu, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.02028
“While some existing works generally recover pose in an end-to-end manner [16,17], limiting the generalization ability for different scenarios, we consider a more practical approach to utilize the deep feature.”
Metadata
- Authors
- Wang, S., Clark, R., Wen, H., & Trigoni, N.
- Affiliations
- University of Oxford (GB)
- Venue
- 2017 IEEE International Conference on Robotics and Automation (ICRA)
- Publisher
- IEEE
- Published
- 2017-05
- Pages
- 2043-2050
- ISSN
- 1050-4729 (ISSN-L)
- OpenAlex
- W2598706937
- Cited by
- 970
- References
- 29
- Citations by year
- 2026: 63 · 2025: 120 · 2024: 104 · 2023: 123 · 2022: 130 · 2021: 131 · 2020: 125 · 2019: 116 · 2018: 52 · 2017: 6
- FWCI
- 143.15
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Proceedings - IEEE International Conference on Robotics and Automation/Proceedings (publishedVersion)
- arXiv (Cornell University) (publishedVersion)
- Warwick Research Archive Portal (University of Warwick) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Engineering and Physical Sciences Research Council (EP/M019918/1)
- 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
- monocular visual odometry, end-to-end learning, recurrent convolutional neural network, pose estimation, feature representation learning
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
