MuST-Cinema
收藏资源简介:
MuST-Cinema是一个多语言语音翻译语料库,由Bruno Kessler基金会创建,包含音频、转录和翻译三元组,特别标注了字幕断点。该数据集解决了现有字幕语料库缺乏源语言音频对齐和字幕断点信息的问题,适用于训练神经机器翻译系统,以自动分割句子为字幕序列。MuST-Cinema的应用领域包括提高字幕翻译的效率和质量,减少人工工作量和相关成本。
MuST-Cinema is a multilingual speech translation corpus developed by the Bruno Kessler Foundation. It consists of audio, transcription and translation triplets, with dedicated annotations for subtitle breakpoints. This corpus addresses a critical limitation of existing subtitle corpora, which lack alignment with source-language audio and subtitle breakpoint information. It is suitable for training neural machine translation systems to automatically segment sentences into subtitle sequences. Its application scenarios include improving the efficiency and quality of subtitle translation, as well as reducing manual workload and associated costs.

- 1MuST-Cinema: a Speech-to-Subtitles corpusBruno Kessler基金会 · 2020年



