The Experts below are selected from a list of 6777 Experts worldwide ranked by ideXlab platform
Felix Siebert - One of the best experts on this subject based on the ideXlab platform.
-
Helmet Use detection of tracked motorcycles using cnn based multi task learning
IEEE Access, 2020Co-Authors: Hanhe Lin, Deike Albers, Jeremiah D Deng, Felix SiebertAbstract:Automated detection of motorcycle Helmet Use through video surveillance can facilitate efficient education and enforcement campaigns that increase road safety. However, existing detection approaches have a number of shortcomings, such as the inabilities to track individual motorcycles through multiple frames, or to distinguish drivers from passengers in Helmet Use. Furthermore, datasets Used to develop approaches are limited in terms of traffic environments and traffic density variations. In this paper, we propose a CNN-based multi-task learning (MTL) method for identifying and tracking individual motorcycles, and register rider specific Helmet Use. We further release the Helmet dataset, which includes 91,000 annotated frames of 10,006 individual motorcycles from 12 observation sites in Myanmar. Along with the dataset, we introduce an evaluation metric for Helmet Use and rider detection accuracy, which can be Used as a benchmark for evaluating future detection approaches. We show that the Use of MTL for concurrent visual similarity learning and Helmet Use classification improves the efficiency of our approach compared to earlier studies, allowing a processing speed of more than 8 FPS on consumer hardware, and a weighted average F-measure of 67.3% for detecting the number of riders and Helmet Use of tracked motorcycles. Our work demonstrates the capability of deep learning as a highly accurate and resource efficient approach to collect critical road safety related data.
-
detecting motorcycle Helmet Use with deep learning
Accident Analysis & Prevention, 2020Co-Authors: Felix Siebert, Hanhe LinAbstract:The continuous motorization of traffic has led to a sustained increase in the global number of road related fatalities and injuries. To counter this, governments are focusing on enforcing safe and law-abiding behavior in traffic. However, especially in developing countries where the motorcycle is the main form of transportation, there is a lack of comprehensive data on the safety-critical behavioral metric of motorcycle Helmet Use. This lack of data prohibits targeted enforcement and education campaigns which are crucial for injury prevention. Hence, we have developed an algorithm for the automated registration of motorcycle Helmet usage from video data, using a deep learning approach. Based on 91,000 annotated frames of video data, collected at multiple observation sites in 7 cities across the country of Myanmar, we trained our algorithm to detect active motorcycles, the number and position of riders on the motorcycle, as well as their Helmet Use. An analysis of the algorithm's accuracy on an annotated test data set, and a comparison to available human-registered Helmet Use data reveals a high accuracy of our approach. Our algorithm registers motorcycle Helmet Use rates with an accuracy of -4.4% and +2.1% in comparison to a human observer, with minimal training for individual observation sites. Without observation site specific training, the accuracy of Helmet Use detection decreases slightly, depending on a number of factors. Our approach can be implemented in existing roadside traffic surveillance infrastructure and can facilitate targeted data-driven injury prevention campaigns with real-time speed. Implications of the proposed method, as well as measures that can further improve detection accuracy are discussed.
-
patterns of motorcycle Helmet Use a naturalistic observation study in myanmar
Accident Analysis & Prevention, 2019Co-Authors: Felix Siebert, Deike Albers, Aung U Naing, Paolo Perego, Chamaiparn SantikarnAbstract:Abstract Developing countries are subject to increased motorization, particularly in the number of motorcycles. As Helmet Use is critical to the safety of motorcycle riders, the goal of this study was to identify observable patterns of Helmet Use, which allow a more accurate assessment of Helmet Use in developing countries. In a video based observation study, 124,784 motorcycle riders were observed at seven observation sites throughout Myanmar. Recorded videos were coded for Helmet Use, number of riders on the motorcycle, rider position, gender, and time of day. Generally, motorcycle Helmet Use in Myanmar was found to be low with only 51.5% percent of riders wearing a Helmet. Helmet Use was highest for drivers (68.1%) and decreased for every additional passenger. It was lowest for children standing on the floorboard of the motorcycle (11.3%). During the day, Helmet Use followed a unimodal distribution, with the highest Use observed during the late morning and lowest Use observed in the early morning and late afternoon. Helmet Use varied significantly between observation sites, ranging from 74.8% in Mandalay to 26.9% in Pakokku. In Mandalay, female riders had a higher Helmet Use than male riders, and Helmet Use decreased drastically on a national holiday in the city. Helmet Use of motorcycle riders in Myanmar follows distinct patterns. Knowledge of these patterns can be Used to design more precise Helmet Use evaluations and guide traffic law policy and police enforcement measures. Video based observation proved to be an efficient tool to collect Helmet Use data.
-
Patterns of motorcycle Helmet Use – a naturalistic observation study in Myanmar
Accident Analysis & Prevention, 2019Co-Authors: Felix Siebert, Deike Albers, U Aung Naing, Paolo Perego, Chamaiparn SantikarnAbstract:Abstract Developing countries are subject to increased motorization, particularly in the number of motorcycles. As Helmet Use is critical to the safety of motorcycle riders, the goal of this study was to identify observable patterns of Helmet Use, which allow a more accurate assessment of Helmet Use in developing countries. In a video based observation study, 124,784 motorcycle riders were observed at seven observation sites throughout Myanmar. Recorded videos were coded for Helmet Use, number of riders on the motorcycle, rider position, gender, and time of day. Generally, motorcycle Helmet Use in Myanmar was found to be low with only 51.5% percent of riders wearing a Helmet. Helmet Use was highest for drivers (68.1%) and decreased for every additional passenger. It was lowest for children standing on the floorboard of the motorcycle (11.3%). During the day, Helmet Use followed a unimodal distribution, with the highest Use observed during the late morning and lowest Use observed in the early morning and late afternoon. Helmet Use varied significantly between observation sites, ranging from 74.8% in Mandalay to 26.9% in Pakokku. In Mandalay, female riders had a higher Helmet Use than male riders, and Helmet Use decreased drastically on a national holiday in the city. Helmet Use of motorcycle riders in Myanmar follows distinct patterns. Knowledge of these patterns can be Used to design more precise Helmet Use evaluations and guide traffic law policy and police enforcement measures. Video based observation proved to be an efficient tool to collect Helmet Use data.
-
pw 1768 assessing motorcycle Helmet Use in developing countries advantages of naturalistic observation over hospital based and road side questionnaire surveys
Injury Prevention, 2018Co-Authors: Felix Siebert, Deike Albers, Paolo Perego, Aye Moe Moe Lwin, Chamaiparn SantikarnAbstract:Although motorcycle Helmets are vital to prevent heavy injuries and fatalities in motorcycle crashes, only one third of low- and middle-income countries (LMIC) regularly collects Helmet Use data. When data is available, it is often undetailed or based on small sample sizes. Hence, methods for regular and detailed monitoring of motorcycle Helmet Use in LMIC are needed. In the light of the application in LMIC, resource-efficiency of these methods has to be considered. Common methods to estimate motorcycle Helmet Use (naturalistic observation, self-reports in questionnaire surveys, hospital based surveys) were compared for their accuracy, practicality, and efficiency. Original empirical data on motorcycle Helmet Use was collected in Myanmar and Tanzania from video-based naturalistic observation, hospital based surveys, and road side questionnaires. This data was further compared to existing data sources for Helmet Use in those countries. In Myanmar, the Helmet Use rate registered in hospitals was significantly lower than the Helmet Use rate observed in the surrounding area. In Tanzania, self-reported Helmet Use in questionnaire surveys was significantly higher than observed Helmet Use. Questionnaire surveys and hospital registration systems were more time consuming and labor-intensive to set up than naturalistic video-based observation. Video-based naturalistic observation is an efficient method to assess Helmet Use and yields more accurate Helmet Use estimates than other methods. Participants’ responses in questionnaire surveys on Helmet Use were found to be biased toward higher Helmet Use numbers, most likely due to social desirability in participants’ responses. Hospital based registration underestimated Helmet Use in Myanmar, as riders were more likely to end up in the hospital without a Helmet. A regular Helmet Use assessment through video based naturalistic observation will allow road safety actors in LMIC to efficiently collect accurate and detailed data of motorcycle Helmet Use.
Brent Edward Hagel - One of the best experts on this subject based on the ideXlab platform.
-
bicycle Helmet Use four years after the introduction of Helmet legislation in alberta canada
Accident Analysis & Prevention, 2011Co-Authors: Mohammad Karkhaneh, Brian H Rowe, L D Saunders, Donald C Voaklander, Brent Edward HagelAbstract:Abstract Background Bicycle Helmets reduce fatal and non-fatal head and face injuries. This study evaluated the effect of mandatory bicycle Helmet legislation targeted at those less than 18 years old on Helmet Use for all ages in Alberta. Methods Two comparable studies were conducted two years before and four years after the introduction of Helmet legislation in Alberta in 2002. Bicyclists were observed in randomly selected sites in Calgary and Edmonton and eight smaller communities from June to October. Helmet wearing and rider characteristics were recorded by trained observers. Poisson regression adjusting for clustering by site was Used to obtain Helmet prevalence (HP) and prevalence ratio (PR) (2006 vs. 2000) estimates. Results There were 4002 bicyclists observed in 2000 and 5365 in 2006. Overall, HP changed from 75% to 92% among children, 30% to 63% among adolescents and 52% to 55% among adults. Controlling for city, location, companionship, neighborhood age proportion Conclusions Bicycle Helmet legislation was associated with a greater increase in Helmet Use among the target age group (
-
Helmet Use and risk of neck injury in skiers and snowboarders
American Journal of Epidemiology, 2010Co-Authors: Brent Edward Hagel, Kelly Russell, Claude Goulet, Alberto Nettelaguirre, Barry I PlessAbstract:In a case-control study, the authors examined the relation between Helmet Use and neck injury among Quebec, Canada, skiers and snowboarders using 10 years of ski patrol data (1995–1996 to 2004–2005). Cases were defined as persons with any neck injury (n = 2,986), an isolated neck injury requiring ambulance evacuation (n = 522), or a cervical spine fracture or dislocation (n = 318). The control group included persons with non-head, non-neck injuries (n = 97,408) in an unmatched analysis. The authors also matched cases with controls injured at the same ski area, during the same activity (skiing vs. snowboarding), and during the same season. Helmet Use was the primary exposure variable. For the unmatched analysis, the authors Used unconditional logistic regression and adjusted for clustering by ski area and other covariates. They Used conditional logistic regression for the matched analysis. Multiple imputation was Used to address missing values. The adjusted odds ratio was 1.09 (95% confidence interval (CI): 0.95, 1.25) for any neck injury, 1.28 (95% CI: 0.96, 1.71) for isolated ambulance-evacuated neck injuries, and 1.02 (95% CI: 0.79, 1.31) for cervical spine fractures or dislocations. Similar results were found in the conditional logistic regression analysis and in analyses restricted to children under age 11 years. These results do not suggest that Helmets increase the risk of neck injuries among skiers and snowboarders.
-
effectiveness of bicycle Helmet legislation to increase Helmet Use a systematic review
Injury Prevention, 2006Co-Authors: Mohammad Karkhaneh, Brent Edward Hagel, Jc Kalenga, Brian H RoweAbstract:Background: Head injuries related to bicycle Use are common and can be serious. They can be prevented or reduced in severity with Helmet Use; however, education has resulted in modest Helmet Use in most developed countries. Helmet legislation has been proposed as a method to increase Helmet wearing; while this social intervention is thought to be effective, no systematic review has been performed. Objectives: This review evaluates the scientific evidence for Helmet Use following legislation to identify the effectiveness of legislative interventions to increase bicycle Helmet Use among all age groups. Search strategy: Comprehensive searches of CENTRAL, MEDLINE, EMBASE, CINAHL, Web of Science, British Education Index, LILACS Database, TRIS (Transport Research Information Service), the grey literature, reference lists, and communication with authors was performed to identify eligible studies. Selection criteria: Eligible studies for this review were community based investigations including cohort studies, controlled before-after studies, interrupted time series studies, non-equivalent control group studies Data collection and analysis: Two reviewers extracted the data regarding the percentage of Helmet Use before and after legislation from each study. Individual and pooled odds ratios were calculated along with 95% confidence intervals. Main results: Out of 86 prescreened articles, 25 were potentially relevant to the topic and 11 were finally included in the review. Of 11 studies, eight were published articles, two were published reports, and one was an unpublished article. One additional survey was incorporated following personal communication with the author. While the baseline rate of Helmet Use among these studies varied between 4% and 59%, after legislation this range changed to 37% and 91%. Helmet wearing proportions increased less than 10% in one study, 10–30% in four studies, and more than 30% in seven studies. While the effectiveness of bicycle Helmet legislation varied (n = 11 studies; OR range: 1.2–22), all studies demonstrated higher proportions of Helmet Use following legislation, particularly when the law was targeted to a specific age group. Conclusions: Legislation increased Helmet Use among cyclists, particularly younger age groups and those with low pre-intervention Helmet wearing proportions. These results support legislative interventions in populations without Helmet legislation.
-
the effect of Helmet Use on injury severity and crash circumstances in skiers and snowboarders
Accident Analysis & Prevention, 2005Co-Authors: Brent Edward Hagel, Claude Goulet, Barry I Pless, Robert W Platt, Yvonne RobitailleAbstract:The aim of this study was to examine the effect of Helmet Use on non-head-neck injury severity and crash circumstances in skiers and snowboarders. We Used a matched case-control study over the November 2001 to April 2002 winter season. 3295 of 4667 injured skiers and snowboarders reporting to the ski patrol at 19 areas in Quebec with non-head, non-neck injuries agreed to participate. Cases included those evacuated by ambulance, admitted to hospital, with restriction of normal daily activities (NDAs) >6 days, with non-Helmet equipment damage, fast self-reported speed, participating on a more difficult run than usual, and jumping-related injury. Controls were injured participants without severe injuries or high-energy crash circumstances and were matched to cases on ski area, activity, day, age, and sex. Conditional logistic regression was Used to relate each outcome to Helmet Use. There was no evidence that Helmet Use increased the risk of severe injury or high-energy crash circumstances. The results suggest that Helmet Use in skiing and snowboarding is not associated with riskier activities that lead to non-head-neck injuries.
Deike Albers - One of the best experts on this subject based on the ideXlab platform.
-
Helmet Use detection of tracked motorcycles using cnn based multi task learning
IEEE Access, 2020Co-Authors: Hanhe Lin, Deike Albers, Jeremiah D Deng, Felix SiebertAbstract:Automated detection of motorcycle Helmet Use through video surveillance can facilitate efficient education and enforcement campaigns that increase road safety. However, existing detection approaches have a number of shortcomings, such as the inabilities to track individual motorcycles through multiple frames, or to distinguish drivers from passengers in Helmet Use. Furthermore, datasets Used to develop approaches are limited in terms of traffic environments and traffic density variations. In this paper, we propose a CNN-based multi-task learning (MTL) method for identifying and tracking individual motorcycles, and register rider specific Helmet Use. We further release the Helmet dataset, which includes 91,000 annotated frames of 10,006 individual motorcycles from 12 observation sites in Myanmar. Along with the dataset, we introduce an evaluation metric for Helmet Use and rider detection accuracy, which can be Used as a benchmark for evaluating future detection approaches. We show that the Use of MTL for concurrent visual similarity learning and Helmet Use classification improves the efficiency of our approach compared to earlier studies, allowing a processing speed of more than 8 FPS on consumer hardware, and a weighted average F-measure of 67.3% for detecting the number of riders and Helmet Use of tracked motorcycles. Our work demonstrates the capability of deep learning as a highly accurate and resource efficient approach to collect critical road safety related data.
-
patterns of motorcycle Helmet Use a naturalistic observation study in myanmar
Accident Analysis & Prevention, 2019Co-Authors: Felix Siebert, Deike Albers, Aung U Naing, Paolo Perego, Chamaiparn SantikarnAbstract:Abstract Developing countries are subject to increased motorization, particularly in the number of motorcycles. As Helmet Use is critical to the safety of motorcycle riders, the goal of this study was to identify observable patterns of Helmet Use, which allow a more accurate assessment of Helmet Use in developing countries. In a video based observation study, 124,784 motorcycle riders were observed at seven observation sites throughout Myanmar. Recorded videos were coded for Helmet Use, number of riders on the motorcycle, rider position, gender, and time of day. Generally, motorcycle Helmet Use in Myanmar was found to be low with only 51.5% percent of riders wearing a Helmet. Helmet Use was highest for drivers (68.1%) and decreased for every additional passenger. It was lowest for children standing on the floorboard of the motorcycle (11.3%). During the day, Helmet Use followed a unimodal distribution, with the highest Use observed during the late morning and lowest Use observed in the early morning and late afternoon. Helmet Use varied significantly between observation sites, ranging from 74.8% in Mandalay to 26.9% in Pakokku. In Mandalay, female riders had a higher Helmet Use than male riders, and Helmet Use decreased drastically on a national holiday in the city. Helmet Use of motorcycle riders in Myanmar follows distinct patterns. Knowledge of these patterns can be Used to design more precise Helmet Use evaluations and guide traffic law policy and police enforcement measures. Video based observation proved to be an efficient tool to collect Helmet Use data.
-
Patterns of motorcycle Helmet Use – a naturalistic observation study in Myanmar
Accident Analysis & Prevention, 2019Co-Authors: Felix Siebert, Deike Albers, U Aung Naing, Paolo Perego, Chamaiparn SantikarnAbstract:Abstract Developing countries are subject to increased motorization, particularly in the number of motorcycles. As Helmet Use is critical to the safety of motorcycle riders, the goal of this study was to identify observable patterns of Helmet Use, which allow a more accurate assessment of Helmet Use in developing countries. In a video based observation study, 124,784 motorcycle riders were observed at seven observation sites throughout Myanmar. Recorded videos were coded for Helmet Use, number of riders on the motorcycle, rider position, gender, and time of day. Generally, motorcycle Helmet Use in Myanmar was found to be low with only 51.5% percent of riders wearing a Helmet. Helmet Use was highest for drivers (68.1%) and decreased for every additional passenger. It was lowest for children standing on the floorboard of the motorcycle (11.3%). During the day, Helmet Use followed a unimodal distribution, with the highest Use observed during the late morning and lowest Use observed in the early morning and late afternoon. Helmet Use varied significantly between observation sites, ranging from 74.8% in Mandalay to 26.9% in Pakokku. In Mandalay, female riders had a higher Helmet Use than male riders, and Helmet Use decreased drastically on a national holiday in the city. Helmet Use of motorcycle riders in Myanmar follows distinct patterns. Knowledge of these patterns can be Used to design more precise Helmet Use evaluations and guide traffic law policy and police enforcement measures. Video based observation proved to be an efficient tool to collect Helmet Use data.
-
pw 1768 assessing motorcycle Helmet Use in developing countries advantages of naturalistic observation over hospital based and road side questionnaire surveys
Injury Prevention, 2018Co-Authors: Felix Siebert, Deike Albers, Paolo Perego, Aye Moe Moe Lwin, Chamaiparn SantikarnAbstract:Although motorcycle Helmets are vital to prevent heavy injuries and fatalities in motorcycle crashes, only one third of low- and middle-income countries (LMIC) regularly collects Helmet Use data. When data is available, it is often undetailed or based on small sample sizes. Hence, methods for regular and detailed monitoring of motorcycle Helmet Use in LMIC are needed. In the light of the application in LMIC, resource-efficiency of these methods has to be considered. Common methods to estimate motorcycle Helmet Use (naturalistic observation, self-reports in questionnaire surveys, hospital based surveys) were compared for their accuracy, practicality, and efficiency. Original empirical data on motorcycle Helmet Use was collected in Myanmar and Tanzania from video-based naturalistic observation, hospital based surveys, and road side questionnaires. This data was further compared to existing data sources for Helmet Use in those countries. In Myanmar, the Helmet Use rate registered in hospitals was significantly lower than the Helmet Use rate observed in the surrounding area. In Tanzania, self-reported Helmet Use in questionnaire surveys was significantly higher than observed Helmet Use. Questionnaire surveys and hospital registration systems were more time consuming and labor-intensive to set up than naturalistic video-based observation. Video-based naturalistic observation is an efficient method to assess Helmet Use and yields more accurate Helmet Use estimates than other methods. Participants’ responses in questionnaire surveys on Helmet Use were found to be biased toward higher Helmet Use numbers, most likely due to social desirability in participants’ responses. Hospital based registration underestimated Helmet Use in Myanmar, as riders were more likely to end up in the hospital without a Helmet. A regular Helmet Use assessment through video based naturalistic observation will allow road safety actors in LMIC to efficiently collect accurate and detailed data of motorcycle Helmet Use.
Chamaiparn Santikarn - One of the best experts on this subject based on the ideXlab platform.
-
patterns of motorcycle Helmet Use a naturalistic observation study in myanmar
Accident Analysis & Prevention, 2019Co-Authors: Felix Siebert, Deike Albers, Aung U Naing, Paolo Perego, Chamaiparn SantikarnAbstract:Abstract Developing countries are subject to increased motorization, particularly in the number of motorcycles. As Helmet Use is critical to the safety of motorcycle riders, the goal of this study was to identify observable patterns of Helmet Use, which allow a more accurate assessment of Helmet Use in developing countries. In a video based observation study, 124,784 motorcycle riders were observed at seven observation sites throughout Myanmar. Recorded videos were coded for Helmet Use, number of riders on the motorcycle, rider position, gender, and time of day. Generally, motorcycle Helmet Use in Myanmar was found to be low with only 51.5% percent of riders wearing a Helmet. Helmet Use was highest for drivers (68.1%) and decreased for every additional passenger. It was lowest for children standing on the floorboard of the motorcycle (11.3%). During the day, Helmet Use followed a unimodal distribution, with the highest Use observed during the late morning and lowest Use observed in the early morning and late afternoon. Helmet Use varied significantly between observation sites, ranging from 74.8% in Mandalay to 26.9% in Pakokku. In Mandalay, female riders had a higher Helmet Use than male riders, and Helmet Use decreased drastically on a national holiday in the city. Helmet Use of motorcycle riders in Myanmar follows distinct patterns. Knowledge of these patterns can be Used to design more precise Helmet Use evaluations and guide traffic law policy and police enforcement measures. Video based observation proved to be an efficient tool to collect Helmet Use data.
-
Patterns of motorcycle Helmet Use – a naturalistic observation study in Myanmar
Accident Analysis & Prevention, 2019Co-Authors: Felix Siebert, Deike Albers, U Aung Naing, Paolo Perego, Chamaiparn SantikarnAbstract:Abstract Developing countries are subject to increased motorization, particularly in the number of motorcycles. As Helmet Use is critical to the safety of motorcycle riders, the goal of this study was to identify observable patterns of Helmet Use, which allow a more accurate assessment of Helmet Use in developing countries. In a video based observation study, 124,784 motorcycle riders were observed at seven observation sites throughout Myanmar. Recorded videos were coded for Helmet Use, number of riders on the motorcycle, rider position, gender, and time of day. Generally, motorcycle Helmet Use in Myanmar was found to be low with only 51.5% percent of riders wearing a Helmet. Helmet Use was highest for drivers (68.1%) and decreased for every additional passenger. It was lowest for children standing on the floorboard of the motorcycle (11.3%). During the day, Helmet Use followed a unimodal distribution, with the highest Use observed during the late morning and lowest Use observed in the early morning and late afternoon. Helmet Use varied significantly between observation sites, ranging from 74.8% in Mandalay to 26.9% in Pakokku. In Mandalay, female riders had a higher Helmet Use than male riders, and Helmet Use decreased drastically on a national holiday in the city. Helmet Use of motorcycle riders in Myanmar follows distinct patterns. Knowledge of these patterns can be Used to design more precise Helmet Use evaluations and guide traffic law policy and police enforcement measures. Video based observation proved to be an efficient tool to collect Helmet Use data.
-
pw 1768 assessing motorcycle Helmet Use in developing countries advantages of naturalistic observation over hospital based and road side questionnaire surveys
Injury Prevention, 2018Co-Authors: Felix Siebert, Deike Albers, Paolo Perego, Aye Moe Moe Lwin, Chamaiparn SantikarnAbstract:Although motorcycle Helmets are vital to prevent heavy injuries and fatalities in motorcycle crashes, only one third of low- and middle-income countries (LMIC) regularly collects Helmet Use data. When data is available, it is often undetailed or based on small sample sizes. Hence, methods for regular and detailed monitoring of motorcycle Helmet Use in LMIC are needed. In the light of the application in LMIC, resource-efficiency of these methods has to be considered. Common methods to estimate motorcycle Helmet Use (naturalistic observation, self-reports in questionnaire surveys, hospital based surveys) were compared for their accuracy, practicality, and efficiency. Original empirical data on motorcycle Helmet Use was collected in Myanmar and Tanzania from video-based naturalistic observation, hospital based surveys, and road side questionnaires. This data was further compared to existing data sources for Helmet Use in those countries. In Myanmar, the Helmet Use rate registered in hospitals was significantly lower than the Helmet Use rate observed in the surrounding area. In Tanzania, self-reported Helmet Use in questionnaire surveys was significantly higher than observed Helmet Use. Questionnaire surveys and hospital registration systems were more time consuming and labor-intensive to set up than naturalistic video-based observation. Video-based naturalistic observation is an efficient method to assess Helmet Use and yields more accurate Helmet Use estimates than other methods. Participants’ responses in questionnaire surveys on Helmet Use were found to be biased toward higher Helmet Use numbers, most likely due to social desirability in participants’ responses. Hospital based registration underestimated Helmet Use in Myanmar, as riders were more likely to end up in the hospital without a Helmet. A regular Helmet Use assessment through video based naturalistic observation will allow road safety actors in LMIC to efficiently collect accurate and detailed data of motorcycle Helmet Use.
Hanhe Lin - One of the best experts on this subject based on the ideXlab platform.
-
Helmet Use detection of tracked motorcycles using cnn based multi task learning
IEEE Access, 2020Co-Authors: Hanhe Lin, Deike Albers, Jeremiah D Deng, Felix SiebertAbstract:Automated detection of motorcycle Helmet Use through video surveillance can facilitate efficient education and enforcement campaigns that increase road safety. However, existing detection approaches have a number of shortcomings, such as the inabilities to track individual motorcycles through multiple frames, or to distinguish drivers from passengers in Helmet Use. Furthermore, datasets Used to develop approaches are limited in terms of traffic environments and traffic density variations. In this paper, we propose a CNN-based multi-task learning (MTL) method for identifying and tracking individual motorcycles, and register rider specific Helmet Use. We further release the Helmet dataset, which includes 91,000 annotated frames of 10,006 individual motorcycles from 12 observation sites in Myanmar. Along with the dataset, we introduce an evaluation metric for Helmet Use and rider detection accuracy, which can be Used as a benchmark for evaluating future detection approaches. We show that the Use of MTL for concurrent visual similarity learning and Helmet Use classification improves the efficiency of our approach compared to earlier studies, allowing a processing speed of more than 8 FPS on consumer hardware, and a weighted average F-measure of 67.3% for detecting the number of riders and Helmet Use of tracked motorcycles. Our work demonstrates the capability of deep learning as a highly accurate and resource efficient approach to collect critical road safety related data.
-
detecting motorcycle Helmet Use with deep learning
Accident Analysis & Prevention, 2020Co-Authors: Felix Siebert, Hanhe LinAbstract:The continuous motorization of traffic has led to a sustained increase in the global number of road related fatalities and injuries. To counter this, governments are focusing on enforcing safe and law-abiding behavior in traffic. However, especially in developing countries where the motorcycle is the main form of transportation, there is a lack of comprehensive data on the safety-critical behavioral metric of motorcycle Helmet Use. This lack of data prohibits targeted enforcement and education campaigns which are crucial for injury prevention. Hence, we have developed an algorithm for the automated registration of motorcycle Helmet usage from video data, using a deep learning approach. Based on 91,000 annotated frames of video data, collected at multiple observation sites in 7 cities across the country of Myanmar, we trained our algorithm to detect active motorcycles, the number and position of riders on the motorcycle, as well as their Helmet Use. An analysis of the algorithm's accuracy on an annotated test data set, and a comparison to available human-registered Helmet Use data reveals a high accuracy of our approach. Our algorithm registers motorcycle Helmet Use rates with an accuracy of -4.4% and +2.1% in comparison to a human observer, with minimal training for individual observation sites. Without observation site specific training, the accuracy of Helmet Use detection decreases slightly, depending on a number of factors. Our approach can be implemented in existing roadside traffic surveillance infrastructure and can facilitate targeted data-driven injury prevention campaigns with real-time speed. Implications of the proposed method, as well as measures that can further improve detection accuracy are discussed.