raw_data/readme.txt

Dense-Captioning Events in Video
================================
Citation:
@inproceedings{krishna2017dense,
    title={Dense-Captioning Events in Videos},
    author={Krishna, Ranjay and Hata, Kenji and Ren, Frederic and Bernstein, Michael and Li, Fei-Fei, and Niebles, Juan Carlos},
    booktitle={ArXiv},
    year={2017}
}

zip file generated by 
  Ranjay Krishna
  @RanjayKrishna
  [email protected]


Files
================================
train_ids.json - Contains video ids of training set.
val_ids.json - Contains video ids of valing set.
test_ids.json - Contains video ids of testing set.
train.json - Contains timestamped captions for videos in training set
val_1.json - Contains timestamped captions for videos in validation set
val_2.json - A second set of annotations for the same vides in validation set

Format for {train/val/test}_ids.json
================================
[v_4Lu8ECLHvK4, v_2D22fVcAcyo, ...]

Format for {train/val_1/val_2}.json
================================
{"v_4Lu8ECLHvK4": {
        'duration': 114.64,
        'sentences': [
            "A woman is seen speaking to ...",
            "...",
        ],
        'timestamps': [
            [1.15, 64.77], 
            [38.4, 105.47]
        ],
    },
 "v_2D22fVcAcyo": ...,
}