End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation

Jung, Heechul; Choi, Min-Kook; Kwon, Soon; Jung, Woo Young

doi:10.1109/icce.2018.8326129

Cited by 8 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the main purposes of the V2X technology is precisely to support the possible and efficient communication mechanisms between vehicles and pedestrians aimed at limiting accidents, sometimes fatal. Recently, in this direction, the Pedestrian Collision Warning (PCW) has been developed [58,59], which can practice wireless modules included in mobile phones, such as Wi-Fi, Bluetooth, and Near Field Communication (NFC). In the detection system based on wireless modules included in the PCW architecture, for instance, as the Wi-Fi module competes, an active security mechanism for VRU, which uses beacon stuffing to alert pedestrians to imminent danger, has been introduced in [60].…”

Section: Vehicle-to-everythingmentioning

confidence: 99%

An Overview of Vehicular Communications

2019

View full text Add to dashboard Cite

The transport sector is commonly subordinate to several issues, such as traffic congestion and accidents. Despite this, in recent years, it is also evolving with regard to cooperation between vehicles. The fundamental objective of this trend is to increase road safety, attempting to anticipate the circumstances of potential danger. Vehicle-to-Vehicle (V2V), Vehicle-to-Infrastructure (V2I) and Vehicle-to-Everything (V2X) technologies strive to give communication models that can be employed by vehicles in different application contexts. The resulting infrastructure is an ad-hoc mesh network whose nodes are not only vehicles but also all mobile devices equipped with wireless modules. The interaction between the multiple connected entities consists of information exchange through the adoption of suitable communication protocols. The main aim of the review carried out in this paper is to examine and assess the most relevant systems, applications, and communication protocols that will distinguish the future road infrastructures used by vehicles. The results of the investigation reveal the real benefits that technological cooperation can involve in road safety.

show abstract

Section: Vehicle-to-everythingmentioning

confidence: 99%

An Overview of Vehicular Communications

2019

View full text Add to dashboard Cite

show abstract

“…The proposed technology can actively provide assistance to the driver in many ways (e.g., blind spot detection, pedestrian detection, obstacle detection, etc.). An interesting CNN-based approach for ADAS proposed by Jung et al [20] was a pedestrian collision warning system. However, it is just limited to pedestrian detection and warning.…”

Section: Related Workmentioning

confidence: 99%

“…ECRUis an attempt to implement the powerful CNN for driving assistance in a more trustworthy and productive way. In the last few years, a number of researchers [20][21][22] have offered outstanding solutions that set the standard for the proposed model. The proposed architecture is a novel idea to implement a CNN for road scene understanding using the semantic segmentation architecture.…”

Section: Introductionmentioning

confidence: 99%

ECRU: An Encoder-Decoder Based Convolution Neural Network (CNN) for Road-Scene Understanding

Yasrab

2018

J. Imaging

View full text Add to dashboard Cite

This research presents the idea of a novel fully-Convolutional Neural Network (CNN)-based model for probabilistic pixel-wise segmentation, titled Encoder-decoder-based CNN for Road-Scene Understanding (ECRU). Lately, scene understanding has become an evolving research area, and semantic segmentation is the most recent method for visual recognition. Among vision-based smart systems, the driving assistance system turns out to be a much preferred research topic. The proposed model is an encoder-decoder that performs pixel-wise class predictions. The encoder network is composed of a VGG-19 layer model, while the decoder network uses 16 upsampling and deconvolution units. The encoder of the network has a very flexible architecture that can be altered and trained for any size and resolution of images. The decoder network upsamples and maps the low-resolution encoder’s features. Consequently, there is a substantial reduction in the trainable parameters, as the network recycles the encoder’s pooling indices for pixel-wise classification and segmentation. The proposed model is intended to offer a simplified CNN model with less overhead and higher performance. The network is trained and tested on the famous road scenes dataset CamVid and offers outstanding outcomes in comparison to similar early approaches like FCN and VGG16 in terms of performance vs. trainable parameters.

show abstract

“…Semantic segmentation has been used to solve road-related problems such as identifying road boundaries for autonomous vehicles [ 17 ], preventing pedestrian collision [ 18 ], and detecting road surface damages [ 19 ]. Our case of road dust differs from the cases above in that the boundary of the object of interest, i.e., dust cloud, may not be as clear as a rigid object.…”

Section: Introductionmentioning

confidence: 99%

Beyond Conventional Monitoring: A Semantic Segmentation Approach to Quantifying Traffic-Induced Dust on Unsealed Roads

de Silva,

Ranasinghe,

Sounthararajah

et al. 2024

Sensors

View full text Add to dashboard Cite

Road dust is a mixture of fine and coarse particles released into the air due to an external force, such as tire–ground friction or wind, which is harmful to human health when inhaled. Continuous dust emission from the road surfaces is detrimental to the road itself and the road users. Due to this, multiple dust monitoring and control techniques are currently adopted in the world. The current dust monitoring methods require expensive equipment and expertise. This study introduces a novel pragmatic and robust approach to quantifying traffic-induced road dust using a deep learning method called semantic segmentation. Based on the authors’ previous works, the best-performing semantic segmentation machine learning models were selected and used to identify dust in an image pixel-wise. The total number of dust pixels was then correlated with real-world dust measurements obtained from a research-grade dust monitor. Our method shows that semantic segmentation can be adopted to quantify traffic-induced dust reasonably. Over 90% of the predictions from both correlations fall in true positive quadrant, indicating that when dust concentrations are below the threshold, the segmentation can accurately predict them. The results were validated and extended for real-time application. Our code implementation is publicly available.

show abstract

End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation

Cited by 8 publications

References 8 publications

An Overview of Vehicular Communications

An Overview of Vehicular Communications

ECRU: An Encoder-Decoder Based Convolution Neural Network (CNN) for Road-Scene Understanding

Beyond Conventional Monitoring: A Semantic Segmentation Approach to Quantifying Traffic-Induced Dust on Unsealed Roads

Contact Info

Product

Resources

About