DIRL Gaze Dataset
Information
Information
DIRL gaze dataset contains 16,196 images with 1920x1080 pixels resolution. 37 Asian volunteers participated in the dataset collection, in which 36 volunteers are Taiwanese, and one volunteer is a Bengali, in which ten volunteers wear eyeglasses. Three volunteers were collected with and without eyeglasses. In this case, we treat these samples as different volunteers for simplification. 26 of volunteers are males, and 11 are females. The ages ranged from 20 to 41.
The setup of dataset collection
The setup of dataset collection
(a) Covered head poses
(a) Environmental Setup
Number of volunteers and gaze images of each head pose
Number of volunteers and gaze images of each head pose
Comparing with several publicly available datasets
Comparing with several publicly available datasets
Publication
Publication
DRIL gaze dataset is available for non-commercial use. If you use our dataset, please cite our paper:
@article{Hsu:2019:LMC:3339884.3311784,author = {Hsu, Chih-Fan and Wang, Yu-Shuen and Lei, Chin-Laung and Chen, Kuan-Ta},title = {Look at Me: Correcting Eye Gaze in Live Video Communication},journal = {ACM Trans. Multimedia Comput. Commun. Appl.},issue_date = {June 2019},volume = {15},number = {2},month = jun,year = {2019},issn = {1551-6857}pages = {38:1--38:21},articleno = {38},numpages = {21},url = {http://doi.acm.org/10.1145/3311784},doi = {10.1145/3311784},acmid = {3311784},publisher = {ACM},address = {New York, NY, USA},keywords = {Eye contact, convolutional neural network, gaze correction, image processing, live video communication},}Acknowledgment
Acknowledgment
Many thanks Mr. Yu-Cheng Chen for assisting the gaze dataset collection.