SID: Stereo Image Dataset for Autonomous Driving in Adverse Conditions
0
Sign in to get full access
Overview
- This paper presents the Stereo Image Dataset (SID) for autonomous driving in adverse weather conditions.
- The dataset includes stereo image pairs captured in various challenging environments, such as rain, fog, and nighttime.
- The goal is to support the development and evaluation of computer vision algorithms for autonomous driving in adverse conditions.
Plain English Explanation
The researchers created a new dataset called the Stereo Image Dataset (SID) to help self-driving car systems work better in bad weather. The dataset includes pairs of stereo images - two images taken from slightly different angles to mimic human vision - captured in various challenging environments like heavy rain, thick fog, and nighttime. The goal is to give developers of self-driving car computer vision algorithms a way to test and improve their systems so they can work reliably even in difficult weather conditions. By having this diverse dataset, researchers can train and evaluate their perception algorithms to make self-driving cars more capable of handling the real-world challenges they'll face.
Technical Explanation
The Stereo Image Dataset (SID) presented in this paper aims to support the development and evaluation of computer vision algorithms for autonomous driving in adverse weather conditions. The dataset contains stereo image pairs captured in various challenging environments, including heavy rain, thick fog, and nighttime. The stereo image pairs are designed to mimic human binocular vision and provide depth information that can be used by perception algorithms for tasks like object detection and scene understanding.
The dataset was collected using a custom multi-camera rig mounted on a vehicle and driven through a variety of real-world environments. The authors provide detailed annotations, including camera calibration parameters, object bounding boxes, and semantic segmentation maps. This data can be used to train and evaluate computer vision models for tasks such as depth estimation, object detection, and semantic segmentation in adverse driving conditions.
Critical Analysis
The SID dataset represents an important contribution to the field of autonomous driving in adverse weather conditions. By providing a diverse and annotated dataset of stereo image pairs, the authors enable researchers to develop and test their perception algorithms more effectively.
However, the dataset is limited to a specific geographic region and may not capture the full diversity of adverse driving conditions encountered globally. Additionally, the authors acknowledge that the dataset does not include dynamic objects or events, which could be a valuable addition for testing the robustness of computer vision systems.
Further research could explore ways to expand the dataset's coverage, both in terms of geographic location and the types of adverse conditions included. Integrating the dataset with other existing datasets, such as ACDC or SEVD, could also help create a more comprehensive resource for the research community.
Conclusion
The Stereo Image Dataset (SID) presented in this paper is a valuable resource for the development and evaluation of computer vision algorithms for autonomous driving in adverse weather conditions. By providing a diverse dataset of stereo image pairs captured in challenging environments, the authors enable researchers to train and test their perception algorithms more effectively.
This dataset has the potential to contribute to the advancement of self-driving car technology, making it more robust and capable of handling the real-world challenges encountered on the road. As the field of autonomous driving continues to evolve, datasets like SID will play a crucial role in pushing the boundaries of what is possible and ensuring the safety and reliability of these systems.
This summary was produced with help from an AI and may contain inaccuracies - check out the links to read the original source documents!
Related Papers
0
SID: Stereo Image Dataset for Autonomous Driving in Adverse Conditions
Zaid A. El-Shair, Abdalmalek Abu-raddaha, Aaron Cofield, Hisham Alawneh, Mohamed Aladem, Yazan Hamzeh, Samir A. Rawashdeh
Robust perception is critical for autonomous driving, especially under adverse weather and lighting conditions that commonly occur in real-world environments. In this paper, we introduce the Stereo Image Dataset (SID), a large-scale stereo-image dataset that captures a wide spectrum of challenging real-world environmental scenarios. Recorded at a rate of 20 Hz using a ZED stereo camera mounted on a vehicle, SID consists of 27 sequences totaling over 178k stereo image pairs that showcase conditions from clear skies to heavy snow, captured during the day, dusk, and night. The dataset includes detailed sequence-level annotations for weather conditions, time of day, location, and road conditions, along with instances of camera lens soiling, offering a realistic representation of the challenges in autonomous navigation. Our work aims to address a notable gap in research for autonomous driving systems by presenting high-fidelity stereo images essential for the development and testing of advanced perception algorithms. These algorithms support consistent and reliable operation across variable weather and lighting conditions, even when handling challenging situations like lens soiling. SID is publicly available at: https://doi.org/10.7302/esz6-nv83.
Read more7/9/2024
0
A-BDD: Leveraging Data Augmentations for Safe Autonomous Driving in Adverse Weather and Lighting
Felix Assion, Florens Gressner, Nitin Augustine, Jona Klemenc, Ahmed Hammam, Alexandre Krattinger, Holger Trittenbach, Sascha Riemer
High-autonomy vehicle functions rely on machine learning (ML) algorithms to understand the environment. Despite displaying remarkable performance in fair weather scenarios, perception algorithms are heavily affected by adverse weather and lighting conditions. To overcome these difficulties, ML engineers mainly rely on comprehensive real-world datasets. However, the difficulties in real-world data collection for critical areas of the operational design domain (ODD) often means synthetic data is required for perception training and safety validation. Thus, we present A-BDD, a large set of over 60,000 synthetically augmented images based on BDD100K that are equipped with semantic segmentation and bounding box annotations (inherited from the BDD100K dataset). The dataset contains augmented data for rain, fog, overcast and sunglare/shadow with varying intensity levels. We further introduce novel strategies utilizing feature-based image quality metrics like FID and CMMD, which help identify useful augmented and real-world data for ML training and testing. By conducting experiments on A-BDD, we provide evidence that data augmentations can play a pivotal role in closing performance gaps in adverse weather and lighting conditions.
Read more8/13/2024
0
FIReStereo: Forest InfraRed Stereo Dataset for UAS Depth Perception in Visually Degraded Environments
Devansh Dhrafani, Yifei Liu, Andrew Jong, Ukcheol Shin, Yao He, Tyler Harp, Yaoyu Hu, Jean Oh, Sebastian Scherer
Robust depth perception in visually-degraded environments is crucial for autonomous aerial systems. Thermal imaging cameras, which capture infrared radiation, are robust to visual degradation. However, due to lack of a large-scale dataset, the use of thermal cameras for unmanned aerial system (UAS) depth perception has remained largely unexplored. This paper presents a stereo thermal depth perception dataset for autonomous aerial perception applications. The dataset consists of stereo thermal images, LiDAR, IMU and ground truth depth maps captured in urban and forest settings under diverse conditions like day, night, rain, and smoke. We benchmark representative stereo depth estimation algorithms, offering insights into their performance in degraded conditions. Models trained on our dataset generalize well to unseen smoky conditions, highlighting the robustness of stereo thermal imaging for depth perception. We aim for this work to enhance robotic perception in disaster scenarios, allowing for exploration and operations in previously unreachable areas. The dataset and source code are available at https://firestereo.github.io.
Read more9/14/2024
0
Bosch Street Dataset: A Multi-Modal Dataset with Imaging Radar for Automated Driving
Karim Armanious, Maurice Quach, Michael Ulrich, Timo Winterling, Johannes Friesen, Sascha Braun, Daniel Jenet, Yuri Feldman, Eitan Kosman, Philipp Rapp, Volker Fischer, Marc Sons, Lukas Kohns, Daniel Eckstein, Daniela Egbert, Simone Letsch, Corinna Voege, Felix Huttner, Alexander Bartler, Robert Maiwald, Yancong Lin, Ulf Ruegg, Claudius Glaser, Bastian Bischoff, Jascha Freess, Karsten Haug, Kathrin Klee, Holger Caesar
This paper introduces the Bosch street dataset (BSD), a novel multi-modal large-scale dataset aimed at promoting highly automated driving (HAD) and advanced driver-assistance systems (ADAS) research. Unlike existing datasets, BSD offers a unique integration of high-resolution imaging radar, lidar, and camera sensors, providing unprecedented 360-degree coverage to bridge the current gap in high-resolution radar data availability. Spanning urban, rural, and highway environments, BSD enables detailed exploration into radar-based object detection and sensor fusion techniques. The dataset is aimed at facilitating academic and research collaborations between Bosch and current and future partners. This aims to foster joint efforts in developing cutting-edge HAD and ADAS technologies. The paper describes the dataset's key attributes, including its scalability, radar resolution, and labeling methodology. Key offerings also include initial benchmarks for sensor modalities and a development kit tailored for extensive data analysis and performance evaluation, underscoring our commitment to contributing valuable resources to the HAD and ADAS research community.
Read more7/19/2024