datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
nhk-archive-audio-30s
NHK Archives Audio 30s
This is a Japanese speech corpus derived from NHK Archives Audio. Audio from public NHK Archives records was segmented into clips of up to 30 seconds using voice activity detection.
The dataset contains 137,594 accepted clips, totaling 1,068.96 hours. Audio is embedded as 16 kHz mono FLAC. raw_text was transcribed with Whisper large-v3-turbo, and text contains LLM-assisted corrections based on the transcript and available source title and description.
This… See the full description on the dataset page: https://huggingface.co/datasets/KeisukeMiyamoto/nhk-archive-audio-30s.nhk-archive-audio
NHK Archives Audio
NHK Archives Audio is a Japanese audio corpus built from records in the public NHK Archives search service. It contains the audio tracks of archive video and audio records together with titles, descriptions, genres, broadcast metadata, regions, source pages, direct stream URLs, and duration metadata.
The source streams were converted to 16 kHz mono FLAC and embedded directly in Parquet files for use with the Hugging Face Dataset Viewer. This dataset does not… See the full description on the dataset page: https://huggingface.co/datasets/KeisukeMiyamoto/nhk-archive-audio.nhk-archive-meta
NHK Archives Metadata
NHK Archives Metadata is a Japanese metadata dataset for audio and video records from the public NHK Archives search service. It provides titles, genres, durations, NHK Archives page URLs, and direct streaming URLs.
The dataset contains metadata and URLs only. It does not contain audio or video files, transcripts, or copied media content.
Purpose
This dataset is intended for research and applications that use Japanese audio and video metadata… See the full description on the dataset page: https://huggingface.co/datasets/KeisukeMiyamoto/nhk-archive-meta.nhkrecipe-100-anno-1
Dataset Card for NHKRecipe-Anno-100
This dataset provides ingredient state annotations for 100 recipes from the NHKRecipe dataset.
Dataset Description
This dataset provides ingredient state annotations for 100 recipes extracted from NHK-supervised recipes (NHKRecipe).
An ingredient state refers to the condition of an ingredient as it changes throughout the cooking process, and is described in natural language for all ingredients present at the end of each cooking step.… See the full description on the dataset page: https://huggingface.co/datasets/mashi6n/nhkrecipe-100-anno-1.nhk_vocabnhk-datasetnHK9fSmZ9ZKzTpK164nHk1wuyQ