IEEE ICIP’16 Challenge Session on Drowsy Driver Detection
Driver Drowsiness Detection from Video
Registration
Announcements
About
Recent reports have suggested that drowsy driving is one of the main factors in fatal motor vehicle crashes each year. In 2014, the US National Sleep Foundation (NSF) pledged an initiative that seeks to raise public awareness on drowsy driving and asked legislators to have law enforcement, regulations and recommendations on drowsy driving and distraction prevention. Therefore, developing active monitoring systems that help drivers avoid accidents in a timely manner is of utmost importance.
Registration
To register, please download and fill out the Registration form and Dataset License Agreement, and email them to skyand55@gmail.com.
Please note that the signed agreement form should be submitted with the registration form to complete the registration.
Please check out the website http://2016.ieeeicip.org/ChallengeSessions.asp for more details of IEEE ICIP 2016 Challenge Session on Drowsy Driver Detection. We would like to invite researchers and developers to register for this Challenge. For any question, please feel free to contact the session organizers by email at (lai@cs.nthu.edu.tw, msarkis@qti.qualcomm.com, ckchiang@cs.ccu.edu.tw).
Shang-Hong Lai, Michel Sarkis, Chen-Kuo Chiang
Organizers of the IEEE ICIP 2016 Challenge Session on Drowsy Driver Detection
| [Mar 23, 2016] | Due to the lack of submission, IEEE ICIP’16 Challenge Session on Drowsy Driver Detection is canceled. |
| [Mar 4, 2016] | The deadline of detection results submission is extended to March 7, 2016,and the evaluation results on the testing dataset will be sent to the participants on March 9, 2016. |
| [Feb 22, 2016] | Testing dataset is available for download now with the same ftp accounts provided to participants when they registered to the Challenge session. The testing dataset consisting of 70 videos is zipped into the file (Testing_Dataset.zip). Please submit the drowsiness detection results for all videos in the testing datatset via email to skyand55@gmail.com before 11:59pm March 1, 2016 (Pacific Standard Time, USA). |
| [Feb 3, 2016] | Annotation files of the training and evaluation datasets are modified. |
| [Jan 7, 2016] | Training and evaluation dataset are released to participants. |
About
Recent reports have suggested that drowsy driving is one of the main factors in fatal motor vehicle crashes each year. In 2014, the US National Sleep Foundation (NSF) pledged an initiative that seeks to raise public awareness on drowsy driving and asked legislators to have law enforcement, regulations and recommendations on drowsy driving and distraction prevention. Therefore, developing active monitoring systems that help drivers avoid accidents in a timely manner is of utmost importance.
Registration
To register, please download and fill out the Registration form and Dataset License Agreement, and email them to skyand55@gmail.com.
Please note that the signed agreement form should be submitted with the registration form to complete the registration.
Please check out the website http://2016.ieeeicip.org/ChallengeSessions.asp for more details of IEEE ICIP 2016 Challenge Session on Drowsy Driver Detection. We would like to invite researchers and developers to register for this Challenge. For any question, please feel free to contact the session organizers by email at (lai@cs.nthu.edu.tw, msarkis@qti.qualcomm.com, ckchiang@cs.ccu.edu.tw).
Shang-Hong Lai, Michel Sarkis, Chen-Kuo Chiang
Organizers of the IEEE ICIP 2016 Challenge Session on Drowsy Driver Detection
Organizers
| Shang-Hong Lai, National Tsing Hua University, Taiwan |
 lai@cs.nthu.edu.tw |
| Michel Sarkis, Qualcomm Technologies Inc., USA |
| msarkis@qti.qualcomm.com |
| Chen-Kuo Chiang, National Chung Cheng University, Taiwan |
| ckchiang@cs.ccu.edu.tw |
Important Dates
| [Jan 7, 2016] | Training and evaluation datasets available to participants |
| [Feb 22, 2016] | Testing dataset released to participants |
| [Mar 1, 2016] | Participants submit the detection results on testing dataset |
| [Mar 7, 2016] | Accuracies for the detection results for all teams announced |
| [Mar 15, 2016] | Paper submission deadline |
| [Apr 30, 2016] | Paper decision notification |
Datasets
Introduction
This Challenge Special Session uses a driver drowsiness video dataset collected by NTHU Computer Vision Lab. The entire dataset (including training, evaluation, testing dataset) contains 36 subjects of different ethnicities recorded with and without wearing glasses/sunglasses under a variety of simulated driving scenarios, including normal driving, yawning, slow blink rate, falling asleep, burst out laughing, etc., under day and night illumination conditions. The subjects were recorded while sitting on a chair and playing a plain driving game with simulated driving wheel and pedals; meanwhile, they were instructed by an experimenter to perform a series of facial expressions. The total time of the entire dataset is about 9 and a half hours.
The training dataset contains 18 subjects with 5 different scenarios (BareFace,Glasses, Night_BareFace, Night_Glasses, Sunglasses). The sequences for each subject including yawning and slow blink rate with nodding are each recorded for about 1 minute long. The sequences corresponding to two most important scenarios, combination of drowsiness-related symptoms (yawning, nodding, slow blink rate) and combination of non-drowsiness related actions (talking, laughing, looking at both sides), are each recorded about for 1.5 minutes long. The evaluation and testing datasets contain 90 driving videos (from the other 18 subjects) with drowsy and non-drowsy status mixed under different scenarios.
Camera Setting and Video Format
We used an active infrared (IR) illumination to acquire IR videos in the dataset collection. The video resolution is 640x480 and was captured at 30 frames per second in AVI format. The dataset is divided into training, evaluation and testing sets. The training and evaluation datasets will be made available to the participants by January 7th 2016, and the testing dataset will be released later without ground truth. The testing videos are produced by mixing videos of different driving scenarios.
Same scenario, different behavior
Same behavior, different scenarios
Database Access
To access the database, please download and fill out the Registration form and Dataset License Agreement, and email them with the subject "IEEE ICIP 2016 Drowsy Driver Detection Challenge" to skyand55@gmail.com. Please note that the signed agreement form should be submitted with the registration form to complete the registration.
Introduction
This Challenge Special Session uses a driver drowsiness video dataset collected by NTHU Computer Vision Lab. The entire dataset (including training, evaluation, testing dataset) contains 36 subjects of different ethnicities recorded with and without wearing glasses/sunglasses under a variety of simulated driving scenarios, including normal driving, yawning, slow blink rate, falling asleep, burst out laughing, etc., under day and night illumination conditions. The subjects were recorded while sitting on a chair and playing a plain driving game with simulated driving wheel and pedals; meanwhile, they were instructed by an experimenter to perform a series of facial expressions. The total time of the entire dataset is about 9 and a half hours.
The training dataset contains 18 subjects with 5 different scenarios (BareFace,Glasses, Night_BareFace, Night_Glasses, Sunglasses). The sequences for each subject including yawning and slow blink rate with nodding are each recorded for about 1 minute long. The sequences corresponding to two most important scenarios, combination of drowsiness-related symptoms (yawning, nodding, slow blink rate) and combination of non-drowsiness related actions (talking, laughing, looking at both sides), are each recorded about for 1.5 minutes long. The evaluation and testing datasets contain 90 driving videos (from the other 18 subjects) with drowsy and non-drowsy status mixed under different scenarios.
Camera Setting and Video Format
We used an active infrared (IR) illumination to acquire IR videos in the dataset collection. The video resolution is 640x480 and was captured at 30 frames per second in AVI format. The dataset is divided into training, evaluation and testing sets. The training and evaluation datasets will be made available to the participants by January 7th 2016, and the testing dataset will be released later without ground truth. The testing videos are produced by mixing videos of different driving scenarios.
Database Access
To access the database, please download and fill out the Registration form and Dataset License Agreement, and email them with the subject "IEEE ICIP 2016 Drowsy Driver Detection Challenge" to skyand55@gmail.com. Please note that the signed agreement form should be submitted with the registration form to complete the registration.
Evaluation Criterion
Each team participating in this Challenge special session need to submit their driver drowsiness detection results for the evaluation on a testing video dataset provided by the organizers. In addition, they will also need to submit a paper describing their system and performance.
The main evaluation criterion in this Challenge is the accuracy of the drowsiness detection in the testing video dataset. We will compare the submitted drowsiness detection results with the ground truth results based on the some commonly used accuracy measures, such as precision, false alarm rate, F-measure, etc.
Detection result submission deadline:March 1, 2016 March 7, 2016
Paper submission deadline: March 15, 2016
For any queries, please email at: skyand55@gmail.com
Each team participating in this Challenge special session need to submit their driver drowsiness detection results for the evaluation on a testing video dataset provided by the organizers. In addition, they will also need to submit a paper describing their system and performance.
The main evaluation criterion in this Challenge is the accuracy of the drowsiness detection in the testing video dataset. We will compare the submitted drowsiness detection results with the ground truth results based on the some commonly used accuracy measures, such as precision, false alarm rate, F-measure, etc.
Detection result submission deadline:
Paper submission deadline: March 15, 2016
For any queries, please email at: skyand55@gmail.com
Detection Results Submission
Each video in the testing dataset should have one drowsiness detection result saved as a text file with the format the same as the text files for the drowsiness annotation file s for the training and evaluation datasets. For the 70 videos in the testing dataset, the submission should contain 70 detection result files placed under the same path and these text files should be compressed into a zip file.
Format of the result :
Detection File name: [video name]_drowsiness_result.txt, ex: 003_glasses_mixing_drowsiness_result.txt
Detection File format: same as the drowsiness annotation files in the training and evaluation datasets
Paper Submission Instructions
Papers should be formatted according to the IEEE ICIP 2016 Paper Kit available under paper submission on 2016.ieeeicip.org. Papers should be submitted online by 15 March 2016 at 2016.ieeeicip.org under the Challenge Special Session: Drowsy Driver Detection topic.
Each video in the testing dataset should have one drowsiness detection result saved as a text file with the format the same as the text files for the drowsiness annotation file s for the training and evaluation datasets. For the 70 videos in the testing dataset, the submission should contain 70 detection result files placed under the same path and these text files should be compressed into a zip file.
Format of the result :
Detection File name: [video name]_drowsiness_result.txt, ex: 003_glasses_mixing_drowsiness_result.txt
Detection File format: same as the drowsiness annotation files in the training and evaluation datasets
Paper Submission Instructions
Papers should be formatted according to the IEEE ICIP 2016 Paper Kit available under paper submission on 2016.ieeeicip.org. Papers should be submitted online by 15 March 2016 at 2016.ieeeicip.org under the Challenge Special Session: Drowsy Driver Detection topic.
Q & A