Data preprocessing

[PSYGIM]

Hello, I noticed that OGM-GE_CVPR2022 does not mention how the KineticSound dataset was processed. If possible, could you kindly provide the preprocessing code for the KineticSound dataset? I would greatly appreciate it!

[echo0409]

Thank you for your question.

For audio, please refer to https://github.com/GeWu-Lab/OGM-GE_CVPR2022/blob/main/data/KineticSound/process_audio.py.

For video, we just extract 1 frame per second. Then we evenly pick 3 frames for each sample during training.