Visual-Inertial Image-Odometry Network (VIIONet): A Gaussian process regression-based deep architecture proposal for UAV pose estimation

Aslan, Muhammet Fatih; Durdu, Akif; Sabancı, Kadir

doi:10.1016/j.measurement.2022.111030

Cited by 30 publications

(14 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CodeVIO [21] proposes a lightweight, tightly-coupled visualinertia odometry network along with depth network to provide accurate state estimates and dense depth estimates of the surroundings. [22] learns the pose of a drone in a more stable way, by denoising inertial data, adopting Inception-v3 to extract visual feature from two consecutive images, and producing pose via Gaussian Process Regression.…”

Section: B Deep Learning Based Visual-inertial Odometrymentioning

confidence: 99%

EMA-VIO: Deep Visual-Inertial Odometry with External Memory Attention

Tu¹,

Chen²,

Pan³

et al. 2022

Preprint

View full text Add to dashboard Cite

Accurate and robust localization is a fundamental need for mobile agents. Visual-inertial odometry (VIO) algorithms exploit the information from camera and inertial sensors to estimate position and translation. Recent deep learning based VIO models attract attentions as they provide pose information in a data-driven way, without the need of designing hand-crafted algorithms. Existing learning based VIO models rely on recurrent models to fuse multimodal data and process sensor signal, which are hard to train and not efficient enough. We propose a novel learning based VIO framework with external memory attention that effectively and efficiently combines visual and inertial features for states estimation. Our proposed model is able to estimate pose accurately and robustly, even in challenging scenarios, e.g., on overcast days and water-filled ground , which are difficult for traditional VIO algorithms to extract visual features. Experiments validate that it outperforms both traditional and learning based VIO baselines in different scenes.

show abstract

Section: B Deep Learning Based Visual-inertial Odometrymentioning

confidence: 99%

EMA-VIO: Deep Visual-Inertial Odometry with External Memory Attention

Tu¹,

Chen²,

Pan³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In multi-view geometry-based odometry estimation, such as ORB-SLAM2 [17] and RTAB-Map [18], the front-end consists of sensor calibration, keypoints extraction and matching, and outlier rejection modules. Under ideal conditions, where the environment is static [19] [20] and there are no issues with texture or illumination, the accuracy of odometry estimation is satisfactory [9]. However, real-world applications require high robustness of the odometry estimation system to deal with challenging scenarios.…”

Section: A Geometry-based and Learning-based Odometry Estimationmentioning

confidence: 99%

“…1) How to fuse multiple modalities: Under the category of fusion strategies, there are concatenate-based approaches such as VINet [7], Li et al's method [8], VIIONet [9], and HVIOnet [10]. These methods directly concatenate the features from different modalities, which means that the weight of each sensor is constant and equal.…”

Section: Learning-based Multi-modal Fusion For Odometry Estimationmentioning

confidence: 99%

“…Before feeding the data into the Trans-FusionOdom framework, we conduct denoising of the IMU signal because M. Brossard et al [44] stated that denoising IMU bias and noise could improve the accuracy of state estimation. Similar to VIIONet [9], we apply the Savitzky-Golay filter [45] to filter the IMU high frequency noise.…”

Section: A Lidar and Imu Data Pre-processingmentioning

confidence: 99%

“…Filter-based and optimization-based solutions have been traditionally employed for sensor fusion in robotics. However, the linearization errors in filter-based methods limit the ac- curacy of estimation [9]. For optimization-based solutions, it is difficult to design adaptive weights, such as a covariance matrix, to balance each residual of the sensor [11].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

CertainOdom: Uncertainty Weighted Multi-task Learning Model for LiDAR Odometry Estimation

Sun

Ding

Yoshiyasu

et al. 2022

2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)

View full text Add to dashboard Cite

Multi-modal fusion of sensors is a commonly used approach to enhance the performance of odometry estimation, which is also a fundamental module for mobile robots. However, the question of how to perform fusion among different modalities in a supervised sensor fusion odometry estimation task? is still one of challenging issues remains. Some simple operations, such as element-wise summation and concatenation, are not capable of assigning adaptive attentional weights to incorporate different modalities efficiently, which make it difficult to achieve competitive odometry results. Recently, the Transformer architecture has shown potential for multi-modal fusion tasks, particularly in the domains of vision with language. In this work, we propose an end-to-end supervised Transformer-based LiDAR-Inertial fusion framework (namely TransFusionOdom) for odometry estimation. The multi-attention fusion module demonstrates different fusion approaches for homogeneous and heterogeneous modalities to address the overfitting problem that can arise from blindly increasing the complexity of the model. Additionally, to interpret the learning process of the Transformer-based multi-modal interactions, a general visualization approach is introduced to illustrate the interactions between modalities. Moreover, exhaustive ablation studies evaluate different multi-modal fusion strategies to verify the performance of the proposed fusion strategy. A synthetic multi-modal dataset is made public to validate the generalization ability of the proposed fusion strategy, which also works for other combinations of different modalities. The quantitative and qualitative odometry evaluations on the KITTI dataset verify the proposed TransFusionOdom could achieve superior performance compared with other related works.

show abstract