Driver behavior detection via adaptive spatial attention mechanism

Zhao, Lei; Yang, Fei; Bu, Lingguo; Su, Han; Zhang, Guoxin; Ying, Luo

doi:10.1016/j.aei.2021.101280

Cited by 35 publications

(18 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By collating previous studies (Montero-Odasso et al, 2009 ; Boripuntakul et al, 2014 ; Zhao et al, 2021 ), it can be seen that most fNIRS studies focus on the prefrontal lobe. The main characteristic indexes discussed are the changes and curves of HbO and HbR in the prefrontal lobe.…”

Section: Methodsmentioning

confidence: 99%

Assessment of Brain Function in Patients With Cognitive Impairment Based on fNIRS and Gait Analysis

Wang

Ren²,

Li³

et al. 2022

Front. Aging Neurosci.

View full text Add to dashboard Cite

BackgroundEarly detection of mild cognitive impairment is crucial in the prevention of Alzheimer’s disease (AD). This study aims to explore the changes in gait and brain co-functional connectivity between cognitively healthy and cognitively impaired groups under dual-task walking through the functional near-infrared spectroscopy (fNIRS) and gait analysis devices.MethodThis study used fNIRS device and gait analysis devices to collect the data of 54 older adults. According to the Mini-mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) scales, the older adults were cognitively healthy (control group) and cognitively impaired (experimental group), of which 38 were in the control group and 16 were in the experimental group. The experiment was divided into a total of three sets of task experiments: a walking-only experiment, a dual-task walking-easy (DTW-easy) experiment, and a dual-task walking-difficult (DTW-difficult) experiment.Main ResultFor the cognitively impaired and cognitively healthy populations, there were no significant differences in overall functional connectivity, region of interest (ROI) connection strength, and gait performance during single-task walking between the two groups.Whereas the performances of DTW differed significantly from the single-task walking in terms of between-group variability of functional connectivity strength change values, and ROI connection strength change values in relation to the dual-task cost of gait. Finally, the cognitively impaired group was significantly more affected by DTW-difficult tasks than the cognitively healthy group.ConclusionThis study provides a new approach to assist in the diagnosis of people with cognitive impairment and provides a new research pathway for the identification of cognitive impairment.

show abstract

Section: Methodsmentioning

confidence: 99%

Assessment of Brain Function in Patients With Cognitive Impairment Based on fNIRS and Gait Analysis

Wang

Ren²,

Li³

et al. 2022

Front. Aging Neurosci.

View full text Add to dashboard Cite

show abstract

“…The proposed improved model is compared with SOTA methods by employing the data used in our study. Compared with the six methods proposed by Eraqi [30], Mase [40], Mase [41] and Zhao [34]. The comparison results are shown in Table 11.…”

Section: G Comparison With Other Methodsmentioning

confidence: 99%

“…By using this approach, the model has still achieved good recognition results when faced with driving scenes with low-light images. Zhao et al [34] proposed a distraction detection model based on an adaptive spatial attention mechanism. The model extracted images through adaptive discriminative space and cropped them through three sub-networks in turn.…”

Section: B Deep Learning Based Detectionmentioning

confidence: 99%

“…At the same time, the BiFPN module is used to strengthen the exchange of feature information and improve the performance of the driver distraction detection model. [22,27], supervised learning [28,29], deep learning [32,34], and related research subfields such as object detection [18] and human behavior recognition [22,23]. The DDBD can be categorized based on different technology into two categories: traditional machine learning based detection and deep learning based detection.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

AB-DLM: An Improved Deep Learning Model Based on Attention Mechanism and BiFPN for Driver Distraction Behavior Detection

Zhang

Li³

et al. 2022

IEEE Access

View full text Add to dashboard Cite

Driver distraction behavior causes a large number of traffic accidents every year, resulting in economic losses and injuries. Currently, the driver still plays an important role in the driving and control of the vehicle due to the low level of vehicle automation and the immature development of autonomous driving. Therefore, it is vital to research distraction detection for drivers. However, in realistic driving scenarios with uncertain information, they are still some challenges in efficient and accurate driver distraction detection. In this paper, an improved deep learning model based on attention mechanisms and bi-directional feature pyramid networks (BiFPN) is proposed to identify driver distractions. Firstly, an improved data augmentation strategy is introduced to increase the data diversity to enhance the generalization capability of the model. Secondly, the squeeze-and-excitation (SE) attention mechanism layer is used after the C3 module of the original backbone network to enhance the important feature information and suppress the minor feature information. Finally, the BiFPN module is introduced into the neck network to better achieve multi-scale feature fusion without increasing the calculation amount too much. The experimental results show that the method proposed in this paper has an average mean accuracy rate (mAP) of 0.956 on the test set. Compared to the original model the mAP has improved by 13.2%. The detection speed of the model is 71 frames per second, and the memory occupation is 15.9 MB. This method has the advantages of high recognition accuracy, fast detection speed, and small memory occupation of the model, which are important for achieving engineering deployment.

show abstract

“…Tran et al [32] deploy a dual-camera system to detect multiple distracted driving behaviors by capturing body movements and face cues. Zhao et al [33] use adaptive spatial attention mechanism for driver activity detection. More recently, Tan et al [34] design a bidirectional posture-appearance interaction network to exploit RGB-and skeleton data in driver behavior recognition.…”

Section: Related Workmentioning

confidence: 99%

TransDARC: Transformer-based Driver Activity Recognition with Latent Space Feature Calibration

Peng¹,

Roitberg²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Traditional video-based human activity recognition has experienced remarkable progress linked to the rise of deep learning, but this effect was slower as it comes to the downstream task of driver behavior understanding. Understanding the situation inside the vehicle cabin is essential for Advanced Driving Assistant System (ADAS) as it enables identifying distraction, predicting driver's intent and leads to more convenient human-vehicle interaction. At the same time, driver observation systems face substantial obstacles as they need to capture different granularities of driver states, while the complexity of such secondary activities grows with the rising automation and increased driver freedom. Furthermore, a model is rarely deployed under conditions identical to the ones in the training set, as sensor placements and types vary from vehicle to vehicle, constituting a substantial obstacle for real-life deployment of data-driven models. In this work, we present a novel vision-based framework for recognizing secondary driver behaviours based on visual transformers and an additional augmented feature distribution calibration module. This module operates in the latent feature-space enriching and diversifying the training set at feature-level in order to improve generalization to novel data appearances, (e.g., sensor changes) and general feature quality. Our framework consistently leads to better recognition rates, surpassing previous state-of-the-art results of the public Drive&Act benchmark on all granularity levels. Our code will be made publicly available at https: //github.com/KPeng9510/TransDARC.

show abstract

Driver behavior detection via adaptive spatial attention mechanism

Cited by 35 publications

References 19 publications

Assessment of Brain Function in Patients With Cognitive Impairment Based on fNIRS and Gait Analysis

Assessment of Brain Function in Patients With Cognitive Impairment Based on fNIRS and Gait Analysis

AB-DLM: An Improved Deep Learning Model Based on Attention Mechanism and BiFPN for Driver Distraction Behavior Detection

TransDARC: Transformer-based Driver Activity Recognition with Latent Space Feature Calibration

Contact Info

Product

Resources

About