The number of data of grand truth images and RGB images are not the same as each other.

hyf015 / egocentric-gaze-prediction

Code for the paper "Predicting Gaze in Egocentric Video by Learning Task-dependent Attention Transition"

62 stars 18 forks source link

The number of data of grand truth images and RGB images are not the same as each other. #13

Closed kazucmpt closed 5 years ago

kazucmpt commented 5 years ago

We read your dataset_preprocessing.py and understood how to generate ground truth and fixation.txt. But the number of data of grand truth isn't same as that of RGB images.

For example, Alireza_American.avi in GTEA datasets plus has 19824 frames, but dataset_preprocessing.py generate 19847 images.

We think it causes error in gaze_full.py. How did you solve it?

hyf015 commented 5 years ago

Other files have similar problems. Some videos are longer than gaze tracking data and some are shorter. I use the shorter one for reference and manually delete some extra files.

kazucmpt commented 5 years ago

Thank you.