A5030291480- Autonomous Vehicle Technology and Safety
- Human-Automation Interaction and Safety
- Traffic and Road Safety
- EEG and Brain-Computer Interfaces
- Advanced Neural Network Applications
- Traffic Prediction and Management Techniques
- Advanced machining processes and optimization
- Traffic control and management
- Sleep and Work-Related Fatigue
- Video Surveillance and Tracking Methods
- Reliability and Maintenance Optimization
- Emotion and Mood Recognition
- Nonlinear Partial Differential Equations
- Manufacturing Process and Optimization
- Safety Warnings and Signage
- Advanced Measurement and Metrology Techniques
- Probabilistic and Robust Engineering Design
- Machine Fault Diagnosis Techniques
- Vehicle Dynamics and Control Systems
- Engineering Diagnostics and Reliability
- Industrial Vision Systems and Defect Detection
- Advanced Mathematical Physics Problems
- Multi-Criteria Decision Making
- Vehicle emissions and performance
- Advanced Mathematical Modeling in Engineering
Jilin University
2015-2025
Chongqing University
2022-2025
Xi'an University of Architecture and Technology
2024
Xi'an University of Technology
2024
Qujing Normal University
2010-2023
Yunnan Provincial Department of Education
2023
City University of Macau
2023
University of Macau
2023
Shenzhen University
2016-2022
Weatherford College
2022
Effectively detecting pedestrians in various environments would significantly improve driving safety for autonomous vehicles. However, the degrpted visibility and blurred outline appearance of pedestrian images captured during hazy weather strongly limit effectiveness current detection methods. To solve this problem, article presents three novel deep learning approaches based on you only look once. The depth wise separable convolution linear bottleneck skills were used to reduce...
Supervised object detection models based on deep learning technologies cannot perform well in domain shift scenarios where annotated data for training is always insufficient. To this end, adaptation knowledge transfer have emerged to handle the problems. A stepwise adaptive YOLO (S-DAYOLO) framework developed which constructs an auxiliary bridge gap and uses a new (DAYOLO) cross-domain tasks. Different from previous solutions, composed of original source images synthetic that are translated...
In the field of autonomous vehicles (AVs), accurately discerning commander intent and executing linguistic commands within a visual context presents significant challenge. This paper introduces sophisticated encoder-decoder framework, developed to address grounding in AVs. Our Context-Aware Visual Grounding (CAVG) model is an advanced system that integrates five core encoders—Text, Emotion, Image, Context, Cross-Modal—with multimodal decoder. integration enables CAVG adeptly capture...
Driver distraction has been identified as one major cause of unsafe driving. The existing studies on cognitive detection mainly focused high-speed driving situations, but less low-speed traffic in urban This paper presents a method for the driver at stop-controlled intersections and compares its feature subsets classification accuracy with that speed-limited highway. In simulator study, 27 subjects were recruited to participate. is induced by clock task taxes visuospatial working memory....
Distracted driving has been recognized as a major challenge to traffic safety improvement. This article presents novel distraction detection method that is based on new deep network. Unlike traditional methods, the proposed uses both temporal information and spatial of electroencephalography (EEG) signals model inputs. Convolutional techniques gated recurrent units were adopted map relationship between drivers' status EEG in time domain. A simulation experiment was conducted examine...
By detecting the defect location in high-resolution insulator images collected by unmanned aerial vehicle (UAV) various environments, occurrence of power failure can be timely detected and caused economic loss reduced. However, accuracies existing detection methods are greatly limited complex background interference small target detection. To solve this problem, two deep learning based on Faster R-CNN (faster region-based convolutional neural network) proposed paper, namely Exact (exact...
End-to-end approaches are one of the most promising solutions for autonomous vehicles (AVs) decision-making. However, deployment these technologies is usually constrained by high computational burden. To alleviate this problem, we proposed a lightweight transformer-based end-to-end model with risk awareness ability AV Specifically, network depth-wise separable convolution and transformer modules was firstly image semantic extraction from time sequences trajectory data. Then, assessed driving...
In this article, a new dataset, the driver emotion facial expression (DEFE) dataset for drivers’ spontaneous emotions analysis is introduced. The includes recordings from 60 participants during driving. After watching selected video-audio clip to elicit specific emotion, each participant completed driving tasks in same scenario and rated his/her emotional responses processes aspects of dimensional method discrete method. study also conducted classification experiments recognize scales arousal,...
In recent years, deep learning technologies for object detection have made great progress and powered the emergence of state-of-the-art models to address problems. Since domain shift can make detectors unstable or even crash, cross-domain becomes very important design detectors. However, traditional always rely on a large amount reliable ground-truth labelling that is laborious, costly, time-consuming. Although an advanced approach CycleGAN has been proposed tasks, ability reduce divergence...
Human emotions are integral to daily tasks, and driving is now a typical task. Creating multi-modal human emotion dataset in tasks an essential step studies. we conducted three experiments collect multimodal psychological, physiological behavioural for (PPB-Emo). In Experiment I, 27 participants were recruited, the in-depth interview method was employed explore driver's viewpoints on scenarios that induce different emotions. For II, 409 questionnaire survey obtain information induces drivers...
As the most direct way to measure true emotional states of humans, EEG-based emotion recognition has been widely used in affective computing applications. In this paper, we aim propose a novel approach that relies on reduced number EEG electrode channels and at same time overcomes negative impact individual differences achieve high accuracy. According statistical significance results power spectral density (PSD) features obtained from SJTU Emotion Dataset (SEED), six candidate sets are...
Abstract Clustering is an unsupervised learning technology, and it groups information (observations or datasets) according to similarity measures. Developing clustering algorithms a hot topic in recent years, this area develops rapidly with the increasing complexity of data volume datasets. In paper, concept introduced, technologies are analyzed from traditional modern perspectives. First, paper summarizes principles, advantages, disadvantages 20 4 algorithms. Then, core elements presented,...