Datasets

1. COCO-Search18

Predicting Goal-directed Human Attention Using Inverse Reinforcement Learning

COCO-Search18 is a laboratory-quality dataset of goal-directed behavior large enough to train deep-network models. It consists of the eye gaze behavior from 10 people searching for each of 18 target-object categories in 6202 natural-scene images, yielding ~300,000 search fixations.

Yupei Chen, Zhibo Yang, Seoyoung Ahn, Dimitris Samaras, Minh Hoai, and Gregory Zelinsky. COCO-Search18 fixation dataset for predicting goal-directed attention control. Scientific Reports, 11 (1), 1-11, 2021. https://www.nature.com/articles/s41598-021-87715-9

Dataset: MIT/Tübingen Saliency Benchmark

Code: Github

2. COCO-CursorSearch

A dataset of mouse cursor movement from 10 people searching for 18 target-object categories in the same 6202 images in COCO-Search18.

Data collection completed. Will be released soon.

3. COCO-FreeView

A dataset of laboratory-quality free-viewing behavior of 10 people using the same 6202 images in COCO-Search18.

Data collection completed. Will be released soon.