UnDeepVO: Monocular Visual Odometry Through Unsupervised Deep Learning

By Ruihao Li, Sen Wang, Zhiqiang Long, Dongbing Gu. 2017

572 citations recorded by the source corpus.

We propose a novel monocular visual odometry (VO) system called UnDeepVO in this paper. UnDeepVO is able to estimate the 6-DoF pose of a monocular camera and the depth of its view by using deep neural networks. There are two salient features of the proposed UnDeepVo:one is the unsupervised deep learning scheme, and the other is the absolute scale recovery. Specifically, we train UnDeepVoby using stereo image pairs to recover the scale but test it by using consecutive monocular images. Thus, UnDeepVO is a monocular system. The loss function defined for training the networks is based on spatial and temporal dense information. A system overview is shown in Fig. 1. The experiments on KITTI dataset show our UnDeepVO achieves good performance in terms of pose accuracy.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Do Not Have Enough Data? Deep Learning to the Rescue! (2019, methodology): “They also include data noising techniques , such as altering words in the input of self-encoder networks in order to generate a different sentence (Xie et al. 2017; Zolna et al. 2017; Li et al. 2018), or introducing noise on the word-embedding level.”
A Comprehensive Survey of Visual SLAM Algorithms (2022, background): “Research studies into the SLAM algorithms considering dynamic environments are essential to increase the algorithms’ robustness to more realistic situations.”
Deep Learning Sensor Fusion for Autonomous Vehicle Perception and Localization: A Review (2020, unknown): “Li et al. [161] trained the CNN network using stereo images instead of using consecutive monocular images.”
Digging Into Self-Supervised Monocular Depth Estimation (2018, background): “Stereo-based approaches have been extended with semi-supervised data [30, 39], generative adversarial networks [1, 48], additional consistency [50], temporal information [33, 73, 3], and for real-time use [49].”
Differentiable SLAM-net: Learning Particle SLAM for Visual Navigation (2021, background): “, they estimate relative motion between consecutive frames without a global map representation [64; 41].”
Monocular depth estimation based on deep learning: An overview (2020, background): “[112] 2018 Stereo sequences √ Absolute scale recovery Wang et al.”
SuperDepth: Self-Supervised, Super-Resolved Monocular Depth Estimation (2018, methodology): “Disparity Smoothness Loss In order to regularize the disparities in textureless low-image gradient regions, we incorporate an edge-aware term (Equation 4), similar to [6], [13], [23].”
Deep Virtual Stereo Odometry: Leveraging Deep Depth Prediction for Monocular Direct Sparse Odometry (2018, methodology): “We also compare DVSO with DeepVO [41], UnDeepVO [26] and SfMLearner [49] which are deep learning based visual odometry systems trained end-to-end on KITTI.”
Learning Two-View Correspondences and Geometry Using Order-Aware Network (2019, methodology): “With the emergence of deep learning, many works attempted to employ learning-based methods to solve geometric matching tasks, including both dense methods [36, 45, 13, 30] and sparse methods [40, 23, 4, 3, 18, 17].”
Geometric Consistency for Self-Supervised End-to-End Visual Odometry (2018, methodology): “To the best of our knowledge, this is the first approach to unsupervised VO estimation that does not require depth prediction as an auxiliary task, as is usually the case [15, 23, 26].”
Every Pixel Counts ++: Joint Learning of Geometry and Motion with 3D Holistic Understanding (2018, methodology): “Some others relies on monocular videos [6], [10], [11], [52], [53], [54] by incorporating 3D camera pose estimation from a motion network.”