ScreenTalk_JA2ZH-XS

ScreenTalk JA2ZH-XS ScreenTalk JA2ZH-XS is a paired dataset of Japanese speech and Chinese translated text released by DataLabX. It is designed for training and evaluating speech translation (ST) and multilingual speech understanding models. The data consists of spoken dialogue extracted from real-world Japanese movies and TV shows. 📦 Dataset Overview Source Language: Japanese (Audio) Target Language: Simplified Chinese (Text) Number of Samples: 10,000 Total Duration:… See the full description on the dataset page: JA2ZH-XS.

Source: Hugging Face Hub (Itbanque/ScreenTalk_JA2ZH-XS). Metadata imported from the dataset’s Hub tags.