Dataset for Cross-Device and Cross-Subject Consistency Evaluation in Visual Fixation Prediction
- Citation Author(s):
-
Yuli Wu
(Institute of Imaging and Computer Vision, RWTH Aachen University, Aachen, Germany)
Henning Konermann
(Institute of Imaging and Computer Vision, RWTH Aachen University, Aachen, Germany)
Emil Mededovic
(Institute of Imaging and Computer Vision, RWTH Aachen University, Aachen, Germany)
Peter Walter
(Department of Ophthalmology, RWTH Aachen University, Aachen, Germany)
Johannes Stegmaier
(Institute of Imaging and Computer Vision, RWTH Aachen University, Aachen, Germany)
- Submitted by:
- Yuli Wu
- Last updated:
- DOI:
- 10.21227/y4m0-ka14
205 views
- Categories:
- Keywords:
Abstract
This dataset accompanies the paper “Evaluating Cross-Device and Cross-Subject Consistency in Visual Fixation Prediction”. We collected eye gaze data using a 30Hz eye tracker embedded in the Aria Glasses (Meta Platforms, Inc., Menlo Park, CA, USA) on 300 images from the MIT1003 dataset, with each image viewed for 3 seconds by 9 subjects (age range 23-39 years), resulting in a total of 243,000 eye fixations. Besides, we also release the average saliency maps from the subjects' visual fixations. The fixation data is structured to allow for cross-device and cross-subject comparisons, enabling researchers to analyze variations in visual attention across different hardware setups and individual observers.
Instructions:
Random Image Selection from MIT1003
- The file "random300.csv" lists the selected image names from the MIT1003 dataset.
- The order of the images corresponds to the indices of the .npy files.
Visual Fixation Results
- The dataset contains 300 .npy files, organized in folders named "s{n}_results".
- Each .npy file stores 90 fixation coordinates.
Average Saliency Maps
- The "average" folder contains 300 .npy files, each representing an average saliency map.
