NITYMED (Night-Time Yawning-Microsleep-Eyeblink-driver Distraction)
Keywords: Driver Drowsiness Detection, Driver State Monitoring, Yawning, Sleepy eyeblink, distraction, microsleep
Submitted by: Nikos Petrellis
Contact Person: Nikos Petrellis (npetrellis at uop dot gr)
Last Updated: 25/5/2022
License: Creative Commons Attribution
The data set is free to be used provided that one or more of the following papers are cited:
- N. Petrellis, P. Christakos, S. Zogas, P. Mousouliotis, G. Keramidas, N. Voros, C. Antonopoulos, “Challenges Towards Hardware Acceleration of the Deformable Shape Tracking Application”, In the Proceedings of the IEEE VLSI SoC Virtual Conference, 4-8 October 2021, Singapore
- N. Petrellis, S. Zogas, P. Christakos, P. Mousouliotis, G. Keramidas, , N. Voros, C. Antonopoulos, “Software Acceleration of the Deformable Shape Tracking Application”, In Proceedings of the ACM, 2nd Symposium on Pattern Recognition and Applications as workshop of ESSE 2021, Larissa, Nov 6-8, 2021.
- N. Petrellis, S. Zogas, P. Christakos, G. Keramidas, P. Mousouliotis, N. Voros, C. Antonopoulos, “High Speed Implementation of the Deformable Shape Tracking Face Alignment Algorithm”, Euromicro DSD 2021, September 1 – 3, 2021, Palermo | Italy, pp. 174-177.
130 videos are available in this dataset displaying drivers in real cars, moving under nighttime conditions where drowsiness detection is more important. The videos have been captured by an in-car camera mounted on the dash of the car. The participating drivers are: 11 males and 10 females with different features (hair color, beard, glasses, etc). The videos are split in 2 categories:
- Yawning: the drivers yawn 3 times in each video lasting approximately 15-25 seconds (107 videos)
- Microsleep: the drivers talk, look around and have microsleeps in videos lasting approximately 2 minutes (21 videos).
This dataset can be used to test and compare algorithms and models for drowsiness detection under nighttime conditions. Other face, mouth and eye tracking applications can also be tested using this dataset. The illumination is natural with a slight boost by the lowest interior car lights in order to simulate the lighting conditions in an avenue since most of the videos were captured on a dark, not crowded road for safety reasons.
All videos are mp4, 25 frames/sec, are mute and are offered in two resolutions:
- HDTV720: 1280 (width)X720 (height), total dataset size: ~700MB
- FULL: 1920 (width)X1080 (height), total dataset size: ~1.6GB
Videos can be used for research and distinct frames from these videos can also be used in research publications. Please click the Access Dataset button below and in the next page give your name, affiliation and your business/academic email. The structure of these directories is the following: