Look, Listen and Recognise
收藏资源简介:
我们提供了一个角色列表,每个剧集的角色名称与对应的演员或女演员名称并列。如果演员或女演员的名称不可用,则会列为未知。CSV文件中包含每个剧集的开始时间、结束时间、说话者和转录文本。开始时间和结束时间以秒为单位。为了评估说话者分割性能,我们还把CSV文件转换为RTTM格式。此外,我们还提供了用于实验的音频文件。
We provide a list of characters, with the name of each character in every episode paired with the corresponding actor or actress. If the name of the actor or actress is unavailable, it is listed as unknown. The CSV file includes the start time, end time, speaker, and transcribed text for each episode. The start and end times are measured in seconds. To evaluate the performance of speaker segmentation, we have also converted the CSV file into RTTM format. Additionally, we have provided audio files for experimental purposes.
数据集概述
数据集名称
- Look, Listen and Recognise : character-aware audio-visual subtitling (ICASSP 2024)
数据集内容
-
Castlist
- 提供每集的演员表,包含角色名和对应演员名,如演员名未知则标记为unknown。
-
CSV文件
- 包含每集的开始时间、结束时间、说话者和转录文本。时间以秒为单位。
-
RTTM格式
- 将CSV文件转换为RTTM格式,用于评估说话者分割性能。
-
音频和视频文件
数据集引用
@article{korbar2023look, author = {Korbar, Bruno and Huh, Jaesung and Zisserman, Andrew}, title = {Look, Listen and Recognise: character-aware audio-visual subtitling}, booktitle = {ICASSP}, year = {2024}, }




