Datasets:
audio dict | id string | sentence_id string | text string | duration_s float64 | sample_rate int64 | bit_depth int64 | audio_format string | categories list | target_terms list | prompt_id string | prompt_source string | mode string | speaker string | speaker_id string | speaker_gender string | contributor string | recorded_at string | device string | microphone string | environment string | transcript_status string | pronunciation string | task string | noise_flags list | model_hypothesis string | asr_model string | notes null | extra string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
{"bytes":"ZkxhQwAAACISABIAAAAAACQVC7gA8AAAAAAAAAAAAAAAAAAAAAAAAAAABAAALAwAAABMYXZmNjIuMy4xMDABAAAAFA(...TRUNCATED) | 20261112T161250Z_ce058f | p_02f106c45a | "And recording a quick note here. And we're trying to do voice isolation. There is a man called Mike(...TRUNCATED) | 183.456 | 48,000 | 16 | flac | [
"background-speech",
"speaker-isolation"
] | [] | drive-voice-note | free-speech | Daniel Rosehill | 8992 | male | Daniel Rosehill | 2026-11-12T16:12:50Z | OnePlus Nord 3 5G | phone built-in | home living room, others present | pending | [] | background-conversations | [
"background-speech",
"baby-crying",
"device-voice-prompt"
] | "And recording a quick note here. And we're trying to do voice isolation. There is a man called Mike(...TRUNCATED) | assemblyai-universal (speaker_labels on; heard one speaker) | null | "{\"background_sources\": [{\"who\": \"Mike\", \"kind\": \"adult male speech\", \"note\": \"lying on(...TRUNCATED) |
Background Conversations
Audio dataset, standalone. Kept apart from the master dataset and its split-outs. Every audio dataset as one tree, with hours: the audio index. How every dataset fits together: the dataset index.
A small test-only set for target-speaker ASR: one person (the foreground speaker) dictates into a phone while other people talk, a baby cries and household devices make noise in the same room. The transcript is the foreground speaker's words only. Everything else is what a model, or a speaker-isolation front end before it, should leave out.
It is never trained on. It is for checking how a dictation model, or a voice-isolation step, copes with real rooms.
Data
- 1 clip(s), 3.1 min, split
testonly. - Audio: FLAC, 48 kHz, 16-bit, mono, converted from phone voice notes (Opus).
text: foreground speech, verbatim, punctuated and cased.transcript_statusispendinguntil the speaker has checked it against the audio, thenreviewed.model_hypothesis/asr_model: the untouched first-pass transcript the draft started from.noise_flags: what is audible behind the speaker (background-speech,baby-crying,device-voice-prompt, ...).extra(JSON text):background_sources(who or what is in the background),diarized(false: the background speech has no transcript or segment labels yet),source_file.
Columns follow the Hebrish audio format standard (audio.schema.json in this repo). There is no sentences tab: these are
free speech, not read sentences.
Licence
CC-BY-4.0.
- Downloads last month
- 51