Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
6.8
3.9
Journals
Sensors
Special Issues
Machine Learning for Autonomous Driving Perception and Prediction
sensors-logo
Submit to Sensors Review for Sensors Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Machine Learning for Autonomous Driving Perception and Prediction

A special issue of Sensors (ISSN 1424-8220). This special issue belongs to the section "Vehicular Sensing".

Deadline for manuscript submissions: closed (1 July 2023) | Viewed by 12389

Special Issue Editors

Dr. Dan Levi

E-Mail Website
Guest Editor
General Motors Advanced Technical Center, Herzlia 46733, Israel
Interests: computer vision and machine learning
Dr. Oded Bialer

E-Mail Website
Guest Editor
General Motors Advanced Technical Center, Herzlia 46733, Israel
Interests: statistical signal processing; estimation theory; sensor array processing; source localization; machine learning; deep learning
Dr. Shaul Oron

E-Mail Website
Guest Editor
General Motors Advanced Technical Center, Herzlia 46733, Israel
Interests: computer vision; visual tracking under severe view-point change

Special Issue Information

Dear Colleagues,

Robust perception is crucial for the reliable operation of autonomous vehicles (AVs). AV perception tasks must process the information from all the sensors mounted on the vehicle, be they cameras, lidars, radars, ultrasonic, etc., and produce an accurate high-level representation of the vehicle’s surroundings (e.g., the 3D positions of lane markings, cars, and pedestrians). AV predictions further extend perception by animating the future states of this environment. The result of these two modules is used by the AV planner to make safe and comfortable path-planning decisions. Any mistake by the former modules might result in suboptimal decisions at best and fatal outcomes at worst. In the last decade, deep learning has become the quintessential technology in computer vision, improving accuracy in most tasks such as object detection, monocular depth estimation and video understanding. Perception for autonomous driving is no exception, and in fact, deep learning has fueled many of the recent advances in this domain. On top of simply using deep learning for perception and prediction tasks, multiple advances in network architectures, sensor fusion, output representation and domain adaptation are contributing to more robust and positionally accurate results. A particular pain point is the requirement for huge human annotated datasets to cope with the infinite variability of road scenes. Recent advances in photo-realistic simulation, domain adaptation and unsupervised learning methods hold promise for meeting this requirement.

This Special Issue aims to publish original, significant and visionary papers describing scientific methods and technologies that improve the robustness, 3D estimation accuracy, computational complexity and implementation feasibility of algorithms for perception. This Special Issue will provide a broad platform for publishing the many rapid advances that have been achieved in the area of perception and prediction for AVs. In this Special Issue, we would like to focus on understanding what the most promising architectures, training methods and data-leveraging methods for efficient model training are. Submissions of scientific results from experts in academia and industry worldwide are strongly encouraged. Contributions may include, but are not limited to:

  • Novel DNN architectures for perception;
  • Novel algorithms for dynamic scene prediction;
  • Network pruning and optimization;
  • Unsupervised and semi-supervised learning;
  • Cross-modal learning;
  • Sensor fusion;
  • Anomaly detection;
  • Offboard perception for automatic data annotation;
  • Synthetic and augmented scene generation for training and validating perception models;
  • End-to-end perception and prediction;
  • Sensory output optimization for perception;
  • Sensor-placement optimization for perception;
  • Data augmentation methods to cope with variability.

Dr. Dan Levi
Dr. Oded Bialer
Dr. Shaul Oron
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Published Papers (4 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

22 pages, 5355 KiB  
Article
A Multi-Task Fusion Strategy-Based Decision-Making and Planning Method for Autonomous Driving Vehicles
by Weiguo Liu, Zhiyu Xiang, Han Fang, Ke Huo and Zixu Wang
Sensors 2023, 23(16), 7021; https://doi.org/10.3390/s23167021 - 08 Aug 2023
Cited by 1 | Viewed by 1206
Abstract
The autonomous driving technology based on deep reinforcement learning (DRL) has been confirmed as one of the most cutting-edge research fields worldwide. The agent is enabled to achieve the goal of making independent decisions by interacting with the environment and learning driving strategies [...] Read more.
The autonomous driving technology based on deep reinforcement learning (DRL) has been confirmed as one of the most cutting-edge research fields worldwide. The agent is enabled to achieve the goal of making independent decisions by interacting with the environment and learning driving strategies based on the feedback from the environment. This technology has been widely used in end-to-end driving tasks. However, this field faces several challenges. First, developing real vehicles is expensive, time-consuming, and risky. To further expedite the testing, verification, and iteration of end-to-end deep reinforcement learning algorithms, a joint simulation development and validation platform was designed and implemented in this study based on VTD–CarSim and the Tensorflow deep learning framework, and research work was conducted based on this platform. Second, sparse reward signals can cause problems (e.g., a low-sample learning rate). It is imperative for the agent to be capable of navigating in an unfamiliar environment and driving safely under a wide variety of weather or lighting conditions. To address the problem of poor generalization ability of the agent to unknown scenarios, a deep deterministic policy gradient (DDPG) decision-making and planning method was proposed in this study in accordance with a multi-task fusion strategy. The main task based on DRL decision-making planning and the auxiliary task based on image semantic segmentation were cross-fused, and part of the network was shared with the main task to reduce the possibility of model overfitting and improve the generalization ability. As indicated by the experimental results, first, the joint simulation development and validation platform built in this study exhibited prominent versatility. Users were enabled to easily substitute any default module with customized algorithms and verify the effectiveness of new functions in enhancing overall performance using other default modules of the platform. Second, the deep reinforcement learning strategy based on multi-task fusion proposed in this study was competitive. Its performance was better than other DRL algorithms in certain tasks, which improved the generalization ability of the vehicle decision-making planning algorithm. Full article
(This article belongs to the Special Issue Machine Learning for Autonomous Driving Perception and Prediction)
►▼ Show Figures

Figure 1

28 pages, 13720 KiB  
Article
Co-Training for Unsupervised Domain Adaptation of Semantic Segmentation Models
by Jose L. Gómez, Gabriel Villalonga and Antonio M. López
Sensors 2023, 23(2), 621; https://doi.org/10.3390/s23020621 - 05 Jan 2023
Cited by 1 | Viewed by 3160
Abstract
Semantic image segmentation is a core task for autonomous driving, which is performed by deep models. Since training these models draws to a curse of human-based image labeling, the use of synthetic images with automatically generated labels together with unlabeled real-world images is [...] Read more.
Semantic image segmentation is a core task for autonomous driving, which is performed by deep models. Since training these models draws to a curse of human-based image labeling, the use of synthetic images with automatically generated labels together with unlabeled real-world images is a promising alternative. This implies addressing an unsupervised domain adaptation (UDA) problem. In this paper, we propose a new co-training procedure for synth-to-real UDA of semantic segmentation models. It performs iterations where the (unlabeled) real-world training images are labeled by intermediate deep models trained with both the (labeled) synthetic images and the real-world ones labeled in previous iterations. More specifically, a self-training stage provides two domain-adapted models and a model collaboration loop allows the mutual improvement of these two models. The final semantic segmentation labels (pseudo-labels) for the real-world images are provided by these two models. The overall procedure treats the deep models as black boxes and drives their collaboration at the level of pseudo-labeled target images, i.e., neither modifying loss functions is required, nor explicit feature alignment. We test our proposal on standard synthetic and real-world datasets for onboard semantic segmentation. Our procedure shows improvements ranging from approximately 13 to 31 mIoU points over baselines. Full article
(This article belongs to the Special Issue Machine Learning for Autonomous Driving Perception and Prediction)
►▼ Show Figures

Figure 1

10 pages, 450 KiB  
Article
Evaluating and Calibrating Uncertainty Prediction in Regression Tasks
by Dan Levi, Liran Gispan, Niv Giladi and Ethan Fetaya
Sensors 2022, 22(15), 5540; https://doi.org/10.3390/s22155540 - 25 Jul 2022
Cited by 29 | Viewed by 3484
Abstract
Predicting not only the target but also an accurate measure of uncertainty is important for many machine learning applications, and in particular, safety-critical ones. In this work, we study the calibration of uncertainty prediction for regression tasks which often arise in real-world systems. [...] Read more.
Predicting not only the target but also an accurate measure of uncertainty is important for many machine learning applications, and in particular, safety-critical ones. In this work, we study the calibration of uncertainty prediction for regression tasks which often arise in real-world systems. We show that the existing definition for the calibration of regression uncertainty has severe limitations in distinguishing informative from non-informative uncertainty predictions. We propose a new definition that escapes this caveat and an evaluation method using a simple histogram-based approach. Our method clusters examples with similar uncertainty prediction and compares the prediction with the empirical uncertainty on these examples. We also propose a simple, scaling-based calibration method that preforms as well as much more complex ones. We show results on both a synthetic, controlled problem and on the object detection bounding-box regression task using the COCO and KITTI datasets. Full article
(This article belongs to the Special Issue Machine Learning for Autonomous Driving Perception and Prediction)
►▼ Show Figures

Figure 1

15 pages, 6674 KiB  
Article
Fast Panoptic Segmentation with Soft Attention Embeddings
by Andra Petrovai and Sergiu Nedevschi
Sensors 2022, 22(3), 783; https://doi.org/10.3390/s22030783 - 20 Jan 2022
Cited by 7 | Viewed by 3838
Abstract
Panoptic segmentation provides a rich 2D environment representation by unifying semantic and instance segmentation. Most current state-of-the-art panoptic segmentation methods are built upon two-stage detectors and are not suitable for real-time applications, such as automated driving, due to their high computational complexity. In [...] Read more.
Panoptic segmentation provides a rich 2D environment representation by unifying semantic and instance segmentation. Most current state-of-the-art panoptic segmentation methods are built upon two-stage detectors and are not suitable for real-time applications, such as automated driving, due to their high computational complexity. In this work, we introduce a novel, fast and accurate single-stage panoptic segmentation network that employs a shared feature extraction backbone and three network heads for object detection, semantic segmentation, instance-level attention masks. Guided by object detections, our new panoptic segmentation head learns instance specific soft attention masks based on spatial embeddings. The semantic masks for stuff classes and soft instance masks for things classes are pixel-wise coherent and can be easily integrated in a panoptic output. The training and inference pipelines are simplified and no post-processing of the panoptic output is necessary. Benefiting from fast inference speed, the network can be deployed in automated vehicles or robotic applications. We perform extensive experiments on COCO and Cityscapes datasets and obtain competitive results in both accuracy and time. On the Cityscapes dataset we achieve 59.7 panoptic quality with an inference speed of more than 10 FPS on high resolution 1024 × 2048 images. Full article
(This article belongs to the Special Issue Machine Learning for Autonomous Driving Perception and Prediction)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-4
Back to TopTop