Data Preparation for EPIC-KITCHENS-100

Download Annotations

You can download the annotation by

bash download_annotation.sh

The annotation should be in data/epic_kitchens-100/annotations/.

• You can also download original csv annotations from official website, and use convert_epic_kitchens_anno.py to generate the annotations that suitable for the codebase.

Download Pre-extracted Features

Please put the downloaded feature under the path: data/epic_kitchens-100/features/.

We provide the following pre-extracted features for EPIC-KITCHENS-100:

Feature	Url	Backbone	Feature Extraction Setting
SlowFast	Google Drive	SlowFast (Epic Finetuned)	30fps, snippet_stride=16, clip_length=32, frame_interval=1
VideoMAE-L	Noun, Verb	VideoMAE-L-16x4x1 (Epic Finetuned)	30fps, snippet_stride=8, clip_length=16, frame_interval=1
InternVideo2	Coming Soon	InternVideo2-1B (Epic Finetuned)	30fps, snippet_stride=8, clip_length=16, frame_interval=1

Download Raw Videos

Please put the downloaded video under the path: data/epic_kitchens-100/raw_data/.

You can download the raw video from official website, then convert the videos into 30 FPS.

If you find difficulty in preparing the data, please email us at shuming.liu@kaust.edu.sa, and we will send you our processed videos.

Citation

@ARTICLE{Damen2022RESCALING,
    title={Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100},
    author={Damen, Dima and Doughty, Hazel and Farinella, Giovanni Maria and Furnari, Antonino 
    and Ma, Jian and Kazakos, Evangelos and Moltisanti, Davide and Munro, Jonathan 
    and Perrett, Toby and Price, Will and Wray, Michael},
    journal   = {International Journal of Computer Vision (IJCV)},
    year      = {2022},
    volume = {130},
    pages = {33–55},
    Url       = {https://doi.org/10.1007/s11263-021-01531-2}
}