About
Audio Description (AD) describes movie content in real time to help visually impaired individuals enjoy movies, where a narration speech briefly summarizes the ongoing plots during pauses in character dialogue, help its audience keep up with the movie.
The AD creation involves extensive work by human experts, which is costly and difficult to cover the vast array of movies and TV shows online. In pursuit of advancing automatic movie narration, Movie101 provides video-aligned AD texts to facilitate research on AI movie understanding, like narration generation and temporal grounding. Find more details in our papers:
Data
Movie
203 movies totaling 353 hours in comedy, romance, action, etc. 10 movies each for validation and testing.
Narration
Video-aligned Chinese and English narration texts, including 71K raw narration segments and 46K merged narrations paragraphs.
Metadata
Movie metadata like introductions, genres, subtitles, character names, and actor portraits. Each movie contains 7.3 characters on average.
Access to Movie101
Annotations: available at our Github Repo
Videos:
Currently you can directly mail to yzihao@ruc.edu.cn with:
1. a signed consent form (both zh&en versions are signed)
2. your name, affiliation, supervisor, and research purpose
After verification
we will support you with the data access as soon as possible
Thank you