Loading...
Loading...
Speech recognition, text-to-speech, speaker identification, music classification, audio event detection
1,909 datasets
21,421 cleaned Georgian speech samples totaling 35 hours were curated by NMikka from Mozilla Common Voice 19.0 in 2026. The collection features 24 kHz mono WAV audio from 12 speakers specifically filtered for speech synthesis and recognition tasks.
Contains audio clips for training a model to recognize the keyword 'Sam'. Each clip is labeled as positive (contains 'Sam') or negative (phonetically similar words). The dataset includes varied speaking styles, speeds, and intonations.
Maxine Annel Pacheco-Ramírez's dataset contains multimodal recordings from a Music-in-Medicine program performance titled 'A Musical Dialogue'. The data includes brain activity, audio, and video, but emotional ratings for participant 5 are missing. It is a large dataset, approximately 7.66 GB in size, and is available under a CC-BY-4.0 license.
A parallel speech corpus containing audio recordings paired with text transcripts for the Gojjam dialect of Amharic. It is curated by leyu-amharic to support speech technology research. The dataset was last updated in March 2026.
An annotated, speaker-relabelled, and loudness-normalised Shona speech dataset prepared through a reproducible Modal-based data engineering pipeline. This release addresses speaker label contamination in the original source labels by replacing identity columns with acoustically-derived speaker assignments. The dataset is authored by manassehzw and was last updated in March 2026.
The dataset shows the number of court sentences issued and requests resolved per municipality in the individual restitution route of Colombia's Land Restitution Unit. It includes columns for hectares ordered for restitution, beneficiary counts, and municipality codes. The data is provided by www.datos.gov.co and was last updated on 2026-03-09.
Nepali language audio data for text-to-speech applications, published on HuggingFace by author lilgoose777. The dataset was last updated on 2026-05-05. Its specific size, format, and content details are not provided in the metadata.
201 FLAC audio files specifically collected for training Automatic Speech Recognition models in the Konkani language. The dataset was uploaded to Hugging Face by alvynabranches and was last updated on March 21, -2026. All audio files are organized within a single directory.
22,196 hours of raw audio from Hong Kong Legislative Council meetings, processed into 20,471 hours of segmented speech. The dataset, created by laubonghaudoi, is split into raw and segmented subsets. It was last updated on 2026-02-26.
Swivuriso is a large-scale multilingual speech dataset targeting over 3000 hours of audio across 7 South African languages. The dataset is developed by dsfsi-anv to support Automatic Speech Recognition and inclusive speech technologies for low-resource African languages. It was last updated on the platform in February 2026.
A September 22, 2006 snapshot of tide gauge station locations along the Massachusetts coastline, sourced from the NOAA Tides and Currents website. The Massachusetts Office of Coastal Zone Management compiled the data, which measures the diurnal tide cycle of two high and low tides per day. The dataset shows where NOAA has placed instruments to monitor sea level changes.
A boundary line defines the landward limit of the Ocean Management Planning Area for Massachusetts, established 0.3 nautical miles from the mean high water shoreline. The data set, provided by SCIOPS, includes official coordinate values and GIS shapefiles for this legal boundary. It was created pursuant to 'An Act Relative to Oceans' to regulate coastal development.
Records from the Massachusetts Shellfish Sanitation Program managed by the Division of Marine Fisheries (MarineFisheries). It details regulatory activities for commercial shellfish harvesting, aquaculture, and local technical partnerships. The dataset originates from the SCIOPS organization via NASA Earthdata.
Road Centerlines is a geospatial dataset representing the centerline of roadways for the City of Bloomington, Indiana, extended to a countywide network. The data includes public roads, named private roads, major multi-use trails, and proposed roadways, with attributes updated from multiple local government sources. The dataset was last updated on March 8, 2026.
Hindi Podcast Asr Dataset is a large-scale collection of raw Hindi podcast audio designed for speech and language model development. It captures real-world interactions across diverse topics and formats. The dataset was created by InfoBayAI and was last updated in March 2026.
A collection of unscripted human monologues in English, spoken by a female voice. The dataset provides 3-minute preview clips intended for use in automatic speech recognition and voice activity detection tasks. The source, author, and specific collection details are not provided.
A digital geologic-GIS map of the Point Lookout Quadrangle in Colorado, composed of GIS data layers and tables. The dataset was produced by the National Park Service's Geologic Resources Inventory program, adapted from a 1999 source map by Griffitts. It is available in multiple GIS formats including a file geodatabase and an OGC geopackage.
A digital geologic-GIS dataset for the Moccasin Quadrangle in Colorado, adapted from a 1999 National Park Service geologic map. The dataset is composed of GIS data layers and tables, available in multiple formats including a file geodatabase and an OGC geopackage. It was completed as part of the NPS Geologic Resources Inventory program and includes ancillary documents with geologic unit descriptions.
A National Park Service Geologic Resources Inventory digital map of the Trail Canyon Quadrangle in Colorado, adapted from a 1999 geologic map by Griffitts. The dataset includes GIS data layers and tables in multiple formats, such as a file geodatabase and geopackage, along with supporting documentation. It was produced by the NPS Geologic Resources Division as part of the Inventory and Monitoring program.
A digital geologic-GIS map of the Wetherill Mesa Quadrangle in Colorado, adapted from a 1999 National Park Service geologic map. The dataset includes GIS data layers and tables available in file geodatabase and geopackage formats, along with ancillary PDF documents containing unit descriptions and metadata. It was produced by the National Park Service's Geologic Resources Inventory program.