DML-iTrack-HDR

About the dataset

DML-iTrack-HDR dataset is an eye tracking dataset for HDR video. Ten HDR video sequences were shown on a Dolby prototype HDR display, and the eye movements of the viewers were recorded by a SensoMotoric Instruments (SMI) eye tracker. These videos were watched by 18 individuals, and fixation density maps (FDM) were obtained for each frame.

 

The Dolby prototype HDR TV display used in our experiment consists of a projector with the resolution of 1024×768 at the back and a 40-inch full HD LCD placed in the front [1]. The contrast range of the projector is 20000:1 and is capable of emitting 15000 ANSI lumens. To deliver HDR viewing experience, the LCD screen is fed by a color stream, which is calibrated with a luminance stream sent to the projector. The maximum brightness level achieved by this HDR display system prototype is 2700 cd/m2.

 

HDR video sequences

The HDR video sequences used in the study are selected from three different sources. Two videos are from the Technicolor HDR dataset [2] used by the Joint Collaborative Team on Video Coding (JCT-VC) [2]. They were created by Technicolor using a simultaneous low exposure/high exposure capture with a rig of two Sony F3 or F56 cameras. The following six sequences were captured at the Stuttgart Media University (HdM) by Froehlich et al.[3], and the final two were captured at the University of British Columbia with RED Scarlet-X cameras, which can capture dynamic range up to 18 stops.

 

content of dataset

This dataset provides the following:

Fixations positions.

Fixation Density Map. A continuous Fixation Density Map (FDM) is calculated by convolving locations of fixation with a Gaussian filter. The sigma of Gaussian is most commonly set to be approximately 1 degree of visual angle.

 

Video Clip	Source	Clip Snapshot	FDM
Balloon 200 frames 30 fps 1920 * 1080	Technicolor[2]		FDM
Market 400 frames 50 fps 1920 * 1080	Technicolor[2]		FDM
Bistro01 151 frames 30 fps 1920 * 1080	Froehlich et al.[3]		FDM
Bistro02 300 frames 25 fps 1920 * 1080	Froehlich et al.[3]		FDM
Bistro03 170 frames 30 fps 1920 * 1080	Froehlich et al.[3]		FDM
Carousel 339 frames 30 fps 1920 * 1080	Froehlich et al.[3]		FDM
Park 439 frames 30 fps 1920 * 1080	Froehlich et al.[3]		FDM
Fishing 371 frames 30 fps 1920 * 1080	Froehlich et al.[3]		FDM
Playground 222 frames 30 fps 2048*1080	DML-HDR		FDM
Mainmall 241 frames 30 fps 2048*1080	DML-HDR		FDM

 

Disclaimer

This data set is free to be used for any non-commercial purposes. Please cite the following papers if you use our data set:

 

E.Nasiopoulos, Y.Dong, and A. Kingstone, "Evaluation of High Dynamic Range Content Viewing Experience Using Eye-Tracking Data," 10th International Conference on Heterogeneous Networking for Quality, Reliability, Security and Robustness, QSHINE, Aug. 2014, Greece.

 

Y.Dong, E.Nasiopoulos, M.T.Pourazad, P.Nasiopoulos, "High Dynamic Range Video Eye Tracking Dataset," 2nd International Conference on Electronics, Signal processing and Communications, ESPCO, Nov 2014, Greece.

 

A. Banitalebi-Dehkordi, Y. Dong, M. T. Pourazad, and Panos Nasiopoulos, “A Learning Based Visual Saliency Fusion Model For High Dynamic Range Video (LBVS-HDR),” 23^rd European Signal Processing Conference, EUSIPCO 2015.

 

Contact

To download the dataset please contact Dr. Panos Nasiopoulos at "panos [at] ece [dot] ubc [dot] ca".

 

reference

[1] H. Seetzen, W. Heidrich, W. Stuerzlinger, G. Ward, L. Whitehead, M. Trentacoste, A. Ghosh and A. Vorozcovs, "High dynamic range display systems," ACM Transactions on Graphics (TOG), vol. 23, no. 3, pp. 760-768, 2004. [2] Y. Olivier, D. Touzé, C. Serre, S. Lasserre, F. L. Léannec and E. François, "Description of HDR sequences," ISO/IEC JTC1/SC29/WG11 MPEG2014/m31957, San Jose, USA, Oct 2014.
[3] J. Froehlich, S. Grandinetti, B. Eberhardt, S. Walter, A. Schilling and H. Brendel, "Creating cinematic wide gamut HDR-video for the evaluation of tone mapping operators and HDR-displays," IS&T/SPIE Electronic Imaging, San Francisco, Feb 2014.
Download: https://hdr-2014.hdm-stuttgart.de/