Object Detection for Connected and Autonomous Vehicles using CNN with Attention Mechanism

Gupta, Abhishek; Illanko, Kandasamy; Fernando, Xavier

doi:10.1109/vtc2022-spring54318.2022.9860676

Cited by 6 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A wide-angle lens, similar to those used in dash cameras, specifically "Wide-angle Navitar MVL4WA lens with focal length 3.5 mm and the effective field of view of 85 • × 62.9" is utilized. The testing dataset in this paper consists of the Tunnel scene described in [23] and available at 1 . Additionally, the evaluation is extended to a dataset captured by a real autonomous vehicle as is described in the Experimental results section.…”

Section: Testing Frameworkmentioning

confidence: 99%

“…An accurate and robust environmental perception system is crucial for the advancement of intelligent transportation, especially in the case of self-driving vehicles [1]. Meeting the requirements of level 5 autonomy, as specified in the J3016 [2] international standard, entails the ability to operate out of the so-called operational design domain.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

How to deal with glare for improved perception of Autonomous Vehicles

Alam,

Kaleem,

Kelouwani

2024

Preprint

View full text Add to dashboard Cite

Vision sensors are versatile and can capture a wide range of visual cues, such as color, texture, shape, and depth. This versatility, along with the relatively inexpensive availability of machine vision cameras, played an important role in adopting vision-based environment perception systems in autonomous vehicles (AVs). However, vision-based perception systems can be easily affected by glare in the presence of a bright source of light, such as the sun or the headlights of the oncoming vehicle at night or simply by light reflecting off snow or ice-covered surfaces; scenarios encountered frequently during driving. In this paper, we investigate various glare reduction techniques, including the proposed saturated pixel-aware glare reduction technique for improved performance of the computer vision (CV) tasks employed by the perception layer of AVs. We evaluate these glare reduction methods based on various performance metrics of the CV algorithms used by the perception layer. Specifically, we considered object detection, object recognition, object tracking, depth estimation, and lane detection which are crucial for autonomous driving. The experimental findings validate the efficacy of the proposed glare reduction approach, showcasing enhanced performance across diverse perception tasks and remarkable resilience against varying levels of glare.keywords-Autonomous vehicles, environment perception, glare reduction, dark channel prior. I. INTRODUCTIONAn accurate and robust environmental perception system is crucial for the advancement of intelligent transportation, especially in the case of self-driving vehicles [1]. Meeting the requirements of level 5 autonomy, as specified in the J3016 [2] international standard, entails the ability to operate out of the so-called operational design domain. Instead of a carefully managed (usually urban) environment with lots of dedicated infrastructure. Autonomous vehicles (AVs) should be able to operate in uncontrollable environments, including challenging weather, glare, haze, and fog causing illumination variation, poorly marked roads, and unpredictable road users [3]. The perception layer in the AV software stack is responsible for timely perceiving the changes happening in the vehicle's environment, through various computer vision (CV) tasks, such as object detection and recognition, depth estimation, lane detection, and more. In recent times, the vision-based perception layer in AVs has gained immense popularity [4]. Several automotive companies, such as Tesla, BMW, and Mobileye have created their vision-based perception systems. This trend can be attributed to several factors, including

show abstract

Section: Testing Frameworkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

How to deal with glare for improved perception of Autonomous Vehicles

Alam,

Kaleem,

Kelouwani

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Tian et al [26] proposed a SA-YOLOv3 detector that strikes a better compromise between detection speed and accuracy. Gupta et al [27] applied both detection and segmentation to the task of road environment object detection to enhance the intelligent adaptive behaviour of self-driving cars. Wang et al [28] propose an autonomous driving detection network for foggy weather that improves the accuracy of object detection in foggy weather scenarios as well as the speed of detection.…”

Section: Introductionmentioning

confidence: 99%

MCS-YOLO: A Multiscale Object Detection Method for Autonomous Driving Road Environment Recognition

Cao

Peng

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Object detection and recognition of road scenes are crucial tasks of the autonomous driving environmental perception system. The low inference speed and accuracy in object detection models hinder the development of autonomous driving technology. Searching for improvement of detection accuracy and speed is still a challenging task. For solving these problems, we proposed an MCS-YOLO algorithm. Firstly, a coordinate attention module is inserted into the backbone to aggregate the feature map's spatial coordinate and cross-channel information. Then, we designed a multiscale small object detection structure to improve the recognition sensitivity of dense small object. Finally, we applied the Swin Transformer structure to the CNN to enable the network to focus on contextual spatial information. Conducting ablation study on the autonomous driving dataset BDD100K, MCS-YOLO algorithm achieves a mean average precision of 53.6% and a recall rate of 48.3%, which are 4.3% and 3.9% better than the YOLOv5s algorithm respectively. In addition, it can achieve real-time detection speed of 55 frames per second in a real scene. The results show that the MCS-YOLO algorithm is effective and superior in the task of automatic driving object detection.

show abstract