Awesome

Malicious or Benign? Towards Effective Content Moderation for Children's Videos [FLAIRS '36]

Malicious or Benign? Towards Effective Content Moderation for Children's Videos
Syed Hammad Ahmed, Muhammad Junaid Khan, H. M. Umer Qaisar, Gita Sukthankar

The dataset frames can be found here.

The dataset annotation labels can be found here.

The Data Loader can be found here.

Dataset Attributes

videoID: YouTube video id

startingTimeStamp: starting timestamp (seconds) of clip relative to the original/published YouTube video

endingTimeStamp: ending timestamp (seconds) of clip relative to the original/published YouTube video

is_anime: the video is an anime or not

is_videoGame: the video is a gameplay of a video-game or not

is_fastRepetitiveMovement: the video contains fast and repetitive motions or not

is_cartoonCharacter: the video contains a cartoon character or not

is_appearanceUnpleasant: the appearance of the cartoon character is unpleasant/disgusting or not

is_violenceActivity: the video containts any violent activity (hitting/destruction/killing) or not

is_obscene: the video containts any obscene/indecent activity or not

is_audio: the video containts any audio

is_loud: the video containts any loud music/noise or not

is_screaming: the video containts any screaming/shouting or not

is_explosion: the video containts any explosion or gunshot sounds or not

videoClass: Ground Truth Label based on video features - Malicious (1) or Benign (0)

audioClass: Ground Truth Label based on audio features - Malicious (1) or Benign (0)

Citation

If you find our work useful in your research, please cite:

 @inproceedings{ahmed2023malicious,
  title={Malicious or Benign? Towards Effective Content Moderation for Children's Videos},
  author={Ahmed, Syed Hammad and Khan, Muhammad Junaid and Qaisar, Hafiz Muhammad Umer and Sukthankar, Gita},
  booktitle={The International FLAIRS Conference Proceedings},
  volume={36},
  year={2023}
}