Neural Network-Based Recent Research Developments in SLAM for Autonomous Ground Vehicles: A Review

Saleem, Hajira; Malekian, Reza; Munir, Hussan

doi:10.1109/jsen.2023.3273913

Cited by 15 publications

(5 citation statements)

References 136 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The work by Saleem et al [29] closely aligns with our survey paper, as these authors delve into the fundamentals of SLAM issues and present numerous deep learning techniques applied to lidar and visual sensors. It is noteworthy, however, that their coverage of radar is confined to potential fusion scenarios with lidar or visual sensors.…”

Section: Related Worksupporting

confidence: 57%

“…The capacity of models to extract keypoints from both point clouds and images has undergone thorough investigation, particularly in the context of VSLAM and lidar-based SLAM methods [8,28,29]. Notably, when compared to other sensors, radar encounters a distinct challenge due to the prevalence of sparse and unreliable data.…”

Section: Odometry Estimationmentioning

confidence: 99%

“…Other optimization methods, such as formulating an appropriate loss, multi-task learning, and optimization learning techniques have been thoroughly investigated for cameras and lidar. The exact architecture and training routine depends on the available data, and on the sensors present in the system [27][28][29].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Brune,

Meisen,

Pomp

2024

Applied Sciences

View full text Add to dashboard Cite

This paper provides an in-depth review of deep learning techniques to address the challenges of odometry and global ego-localization using frequency modulated continuous wave (FMCW) radar sensors. In particular, we focus on the prediction of odometry, which involves the determination of the ego-motion of a system by external sensors, and loop closure detection, which concentrates on the determination of the ego-position typically on an existing map. We initially emphasize the significance of these tasks in the context of radar sensors and underscore the motivations behind them. The subsequent sections delve into the practical implementation of deep learning approaches, strategically designed to effectively address the aforementioned challenges. We primarily focus on spinning and automotive radar configurations within the domain of autonomous driving. Additionally, we introduce publicly available datasets that have been instrumental in addressing these challenges and analyze the importance and struggles of current methods used for radar based odometry and localization. In conclusion, this paper highlights the distinctions between the addressed tasks and other radar perception applications, while also discussing their differences from challenges posed by alternative sensor modalities. The findings contribute to the ongoing discourse on advancing radar sensor capabilities through the application of deep learning methodologies, particularly in the context of enhancing odometry and ego-localization for autonomous driving applications.

show abstract

Section: Related Worksupporting

confidence: 57%

Section: Odometry Estimationmentioning

confidence: 99%

See 1 more Smart Citation

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Brune,

Meisen,

Pomp

2024

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Deep Learning (DL) has received considerable attention for its remarkable success in Computer Vision (CV) [10][11][12] and Robotics [13], mainly due to the widespread use of Convolutional Neural Networks (CNNs). Pre-trained CNN models are usually used as feature extractors.…”

Section: Deep Learning For Vprmentioning

confidence: 99%

BinVPR: Binary Neural Networks towards Real-Valued for Visual Place Recognition

Wang,

Han,

Dong

et al. 2024

Sensors

View full text Add to dashboard Cite

Visual Place Recognition (VPR) aims to determine whether a robot or visual navigation system locates in a previously visited place using visual information. It is an essential technology and challenging problem in computer vision and robotic communities. Recently, numerous works have demonstrated that the performance of Convolutional Neural Network (CNN)-based VPR is superior to that of traditional methods. However, with a huge number of parameters, large memory storage is necessary for these CNN models. It is a great challenge for mobile robot platforms equipped with limited resources. Fortunately, Binary Neural Networks (BNNs) can reduce memory consumption by converting weights and activation values from 32-bit into 1-bit. But current BNNs always suffer from gradients vanishing and a marked drop in accuracy. Therefore, this work proposed a BinVPR model to handle this issue. The solution is twofold. Firstly, a feature restoration strategy was explored to add features into the latter convolutional layers to further solve the gradient-vanishing problem during the training process. Moreover, we identified two principles to address gradient vanishing: restoring basic features and restoring basic features from higher to lower layers. Secondly, considering the marked drop in accuracy results from gradient mismatch during backpropagation, this work optimized the combination of binarized activation and binarized weight functions in the Larq framework, and the best combination was obtained. The performance of BinVPR was validated on public datasets. The experimental results show that it outperforms state-of-the-art BNN-based approaches and full-precision networks of AlexNet and ResNet in terms of both recognition accuracy and model size. It is worth mentioning that BinVPR achieves the same accuracy with only 1% and 4.6% model sizes of AlexNet and ResNet.

show abstract

“…The advent of deep learning technologies has ushered in a paradigm shift in the landscape of multisensor fusion for SLAM and path planning. By harnessing the data-driven power of neural networks, researchers have circumvented some of the most vexing limitations associated with traditional, rule-based algorithms (Saleem et al, 2023;J.L. Yu et al, 2020).…”

Section: Deep Learning-based Fusionmentioning

confidence: 99%

Intelligent Systems in Motion

Cai,

Qin,

et al. 2023

International Journal on Semantic Web and Information Systems

View full text Add to dashboard Cite

Simultaneous localization and mapping (SLAM) serves as a cornerstone in autonomous systems and has seen exponential growth in its roles, particularly in facilitating advanced path planning solutions. One emerging avenue of research that is rapidly evolving is the incorporation of multi-sensor fusion techniques to enhance SLAM-based path planning. The paper initiates with a thorough review of various sensor types and their attributes before covering a broad spectrum of both traditional and contemporary algorithms for multi-sensor fusion within SLAM. Performance evaluation metrics pertinent to SLAM and sensor fusion are explored. A special focus is laid on the interconnected roles and applications of multi-sensor fusion in SLAM-based path planning, discussing its significance in navigation scenarios as well as addressing challenges such as computational burden and real-time implementation. This paper sets the stage for future developments in creating more robust, resilient, and efficient SLAM-based path planning systems enabled by multi-sensor fusion.

show abstract

Neural Network-Based Recent Research Developments in SLAM for Autonomous Ground Vehicles: A Review

Cited by 15 publications

References 136 publications

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

BinVPR: Binary Neural Networks towards Real-Valued for Visual Place Recognition

Intelligent Systems in Motion

Contact Info

Product

Resources

About