9,029 Hits in 5.5 sec

Robust Lane Detection via Expanded Self Attention [article]

Minhyeok Lee, Junhyeop Lee, Dogyoon Lee, Woojin Kim, Sangwon Hwang, Sangyoun Lee
2021 arXiv   pre-print
In this paper, we propose a simple but powerful self-attention mechanism optimized for lane detection called the Expanded Self Attention (ESA) module.  ...  Modern deep learning methods achieve high performance in lane detection, but it is still difficult to accurately detect lanes in challenging situations such as congested roads and extreme lighting conditions  ...  Inspired by the simple geometry of lanes, ESA modules are divided into HESA (Horizontal Expanded Self Attention) and VESA (Vertical Expanded Self Attention).  ... 
arXiv:2102.07037v3 fatcat:bsiorgsm5bfzbk6qspi7xl7rfq

Lane detection with Position Embedding [article]

Jun Xie, Jiacheng Han, Dezhen Qi, Feng Chen, Kaer Huang, Jianwei Shuai
2022 arXiv   pre-print
Recently, lane detection has made great progress in autonomous driving. RESA (REcurrent Feature-Shift Aggregator) is based on image segmentation.  ...  For Tusimple dataset, there is not too complicated scene and lane has more prominent spatial features.  ...  End-to-end lane marker detection via row-wise classification.  ... 
arXiv:2203.12301v1 fatcat:qd4lf3nvczdb5fxtqn3xzeu7je

Lane Line Detection Based on Improved Semantic Segmentation in Complex Road Environment

Chaowei Ma, Dean Luo, He Huang
2021 Sensors and materials  
The Visual Geometry Group-Special Convolutional Neural Network (VGG-SS) proposed in this paper, which is based on the VGG-16 network, introduces a self-attentive distillation model and a spatial convolutional  ...  To address the current challenges in lane line detection, in this study, we propose a lane line detection model based on improved semantic segmentation for complex road scenarios, such as lane line occlusion  ...  Similar to the self-attention mechanism, (30) the SAD allows a network to use the attention map of its own layer as a learning target for its lower layers, and this attention detection mechanism has  ... 
doi:10.18494/sam.2021.3544 fatcat:6tmavd5nhze5vb54miixju52si

PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark [article]

Li Chen, Chonghao Sima, Yang Li, Zehan Zheng, Jiajie Xu, Xiangwei Geng, Hongyang Li, Conghui He, Jianping Shi, Yu Qiao, Junchi Yan
2022 arXiv   pre-print
Methods for 3D lane detection have been recently proposed to address the issue of inaccurate lane layouts in many autonomous driving scenarios (uphill/downhill, bump, etc.).  ...  OpenLane contains 200,000 frames, over 880,000 instance-level lanes, 14 lane categories, along with scene tags and the closed-in-path object annotations to encourage the development of lane detection and  ...  Finally, we try to remove the classical self attention Visualization.  ... 
arXiv:2203.11089v3 fatcat:5fxllq35ybhctkg4oikiyowsdq

K-Lane: Lidar Lane Dataset and Benchmark for Urban Roads and Highways [article]

Donghee Paek, Seung-Hyun Kong, Kevin Tirta Wijaya
2022 arXiv   pre-print
Lane detection is a critical function for autonomous driving.  ...  We also provide baseline networks we term Lidar lane detection networks utilizing global feature correlator (LLDN-GFC).  ...  LLDN-GFC utilizes self-attention mechanisms to extract lane features via global correlation, and show superior performance compared to the conventional CNN-based LLDNs.  ... 
arXiv:2110.11048v2 fatcat:vqh7dr22vffe7egguqjjxhca2i

Exploring Self-Attention for Visual Intersection Classification [article]

Haruki Nakata, Kanji Tanaka, Koji Takeda
2022 arXiv   pre-print
In robot vision, self-attention has recently emerged as a technique for capturing non-local contexts.  ...  First, we proposed a self-attention-based approach for intersection classification using TPVs.  ...  In [9] , a large-scale input image was used for crack detection. In [6] , a monocular detection method was proposed for robust road detection under severe occlusions.  ... 
arXiv:2203.13977v1 fatcat:6s3bcdttvrexfgk3kegmfzessu

Road Sign Recognition Method Based on Segmentation and Attention Mechanism

Tianao Chen, Aotian Chen, Yajuan Tang
2022 Mobile Information Systems  
, solve the imbalance of positive and negative lane line detection samples, and obtain the final lane line detection result via postprocessing.  ...  When only lane line detection is required, the prediction branch of lane line existence is introduced based on the lightweight semantic segmentation model ERFNet to realize lane line instantiation cognition  ...  [14] proposed self-attentional distillation (SAD) for lane detection, and realized further learning by implementing topdown and hierarchical attention distillation networks within the network.  ... 
doi:10.1155/2022/6389580 fatcat:h3xcpfxodvhgfhra2trmi4efvu

Generation and Adaptation of Transferable Roadway Model for Anticipative Road Following on Satellite-Roadway-Vehicle Network

2011 SICE Journal of Control Measurement and System Integration  
First, the chromatic complexity of the roadway area is represented as a palette of saliency colors via fractal sampling of the scene image.  ...  By matching roadway images in an encountered scene with bird's eye views, the scope of humans' perception is expanded to a satellite-roadway-vehicle network.  ...  In this experiment, the random samples were generated via the self-similarity process illustrated in Fig. 4 ; the image plane Ω with four vertexes are reduced into four copies via a system of contraction  ... 
doi:10.9746/jcmsi.4.97 fatcat:kofkevbp7bar7alo3cyxpwnt5u

End-to-End Deep Learning of Lane Detection and Path Prediction for Real-Time Autonomous Driving [article]

Der-Hau Lee, Jinn-Liang Liu
2021 arXiv   pre-print
These results show that DSUNet is efficient and effective for lane detection and path prediction in autonomous driving.  ...  Inspired by the UNet architecture of semantic image segmentation, we propose a lightweight UNet using depthwise separable convolutions (DSUNet) for end-to-end learning of lane detection and path prediction  ...  The [9] Zou, Q., et al.: Robust lane detection from continuous driving scenes using κ and ∆ curves are reciprocal in peaks and valleys as in their deep neural networks  ... 
arXiv:2102.04738v2 fatcat:lso75tzjcrehhdtcd2wt6oxe6q

Guest Editorial: AI Applications to Intelligent Vehicles for Advancing Intelligent Transport Systems

2020 IET Intelligent Transport Systems  
In 'Real-time running detection system for UAV imagery based on optical flow and deep convolutional networks', Wu et al. present a fast running human detection system for unmanned aerial vehicle (UAV)  ...  In 'Spatio-temporal expand-and-squeeze networks for crowd flow prediction in metropolis', Yang et al. propose a novel framework ST-ESNet, spatial-temporal expand-and-squeeze networks, that designs several  ... 
doi:10.1049/iet-its.2020.0189 fatcat:7cgsidr4lfeqvd3ffjtmv4guwi

A survey of video processing techniques for traffic applications

V Kastrinaki, M Zervakis, K Kalaitzakis
2003 Image and Vision Computing  
† Temporal estimation of vehicle's state variables † Spatial signature for object detection via color segment. † Moving camera NAVLAB [11] † Spatial-domain lane finding † Lane-region detection for alf  ...  . [14] † Spatial processing for alf and object detection † Feature-driven approach † Autonomous vehicle guidance † Feature tracking in temporal domain † Color road detection and lane detection via RLS  ...  In one scheme, the estimated lane borders at the previous frame can be expanded, making the lane virtually wider, so that the actual lane borders at the next frame are searched for within this expanded  ... 
doi:10.1016/s0262-8856(03)00004-0 fatcat:nr3mmyj55jaxxljkvst4laivxy

Computer vision in roadway transportation systems: a survey

Robert P. Loce, Edgar A. Bernal, Wencheng Wu, Raja Bala
2013 Journal of Electronic Imaging (JEI)  
increased automation, robustness, and self-sufficiency of incident detection systems.  ...  This module includes two main phases: lane detection and lane tracking.  ... 
doi:10.1117/1.jei.22.4.041121 fatcat:chkul4gryvawrbxhaagxhkrita

Fast Drivable Areas Estimation with Multi-Task Learning for Real-Time Autonomous Driving Assistant

Dong-Gyu Lee
2021 Applied Sciences  
We propose a fast and accurate multi-task learning-based architecture for joint segmentation of drivable area, lane line, and classification of the scene.  ...  The proposed method learns end-to-end through multi-task learning on a very challenging Berkeley Deep Drive dataset and shows its robustness for three tasks in autonomous driving.  ...  Expanded Self Attention (ESA) [13] propose a self-attention mechanism that can predict the confidence of a lane along with the vertical and horizontal directions.  ... 
doi:10.3390/app112210713 fatcat:fap5khim6fd2rmhm5tmmamamdq

Scanning the Issue

Azim Eskandarian
2021 IEEE transactions on intelligent transportation systems (Print)  
Semantic Scene Labeling via Adversarial Confidence Estimation Networks for Robust Stereo Matching S. Kim, D. Min, S. Kim, and K.  ...  A comparison of the performance with many stateof-the-art models illustrates the advantages and robustness of ResLSTM. A Robust Attentional Autonomous Vehicle: Security by Design A.  ...  Lin This article proposes a unified neural network called Attentive Traffic Flow Machine (ATFM), which can effectively learn the spatial-temporal feature representations of traffic flow with an attention  ... 
doi:10.1109/tits.2021.3119785 fatcat:gjykhcqo3bbwdm5oilmqn7t2re

Innovative RNAi Strategies and Tactics to Tackle Plum Pox Virus (PPV) Genome in Prunus domestica-Plum

Ravelonandro, Scorza, Briard
2019 Plants  
Plum trees transformed with amisiCPRNA possess the two major hallmarks of silencing, first the ability to self-amplify ami and siRNA and secondly, to spread via the vascular tissues in the entire plants  ...  The numbers (upper lanes) represent the clones studied. Arrow (right margin) indicates the expected bands detected.  ... 
doi:10.3390/plants8120565 pmid:31810364 pmcid:PMC6963518 fatcat:m7cbq62vvfhznks2ub3wdcno7i
« Previous Showing results 1 — 15 out of 9,029 results