{ "id": "LK-SP-HIN-004", "slug": "hindi-speaker-diarization", "title": "Hindi speaker diarization", "modality": "speech", "speechType": "conversational", "kind": "diarization", "unit": "speaker turns", "collection": "mixed", "language": { "name": "Hindi", "native": "हिन्दी", "iso": "hin", "bcp47": "hi-IN", "script": "Devanagari" }, "languages": [ { "name": "Hindi", "native": "हिन्दी", "iso": "hin", "bcp47": "hi-IN", "script": "Devanagari" } ], "region": "India", "domain": "insurance calls, general conversation", "scripted": null, "parties": null, "hours": 321.2, "transcribed": true, "channels": "mono", "source": "Recorded for the dataset", "review": "Every recording has a complete, segment-level transcript", "personalData": "Redacted", "licence": "Custom", "sheetRows": [ 5, 14 ], "sample": { "key": "diar-hi", "slug": "hindi-speaker-diarization", "dual": false, "audio": { "sampleRateHz": 16000, "bitDepth": 16, "codec": "pcm_s16le", "container": "flac", "filesPerConversation": 1, "measured": true }, "metrics": { "durationS": 1458.6, "conversations": 6, "turns": 320, "words": 3402, "talkTimeS": { "1": 575.3, "2": 777.9 }, "speechShare": 0.833, "overlapShare": 0.094, "silenceShare": 0.167, "wordsPerMinute": 168, "turnDurationS": { "median": 1.94, "p90": 12.58 }, "floorTransfer": { "count": 172, "medianS": 0.537, "p10S": -0.832, "p90S": 2.189, "overlappedShare": 0.192, "histogram": { "edgesS": [ -2, -1, -0.5, -0.2, 0, 0.2, 0.5, 1, 2, 4 ], "counts": [ 14, 3, 7, 4, 5, 21, 31, 42, 24, 17, 4 ] } }, "backchannels": { "count": 69, "perMinute": 2.84 }, "codeMix": { "latinTokens": 309, "nativeTokens": 3093, "latinShare": 0.091 }, "events": { "[silence]": 24, "[filler]": 38, "[unintelligible]": 58 }, "overlapEvents": 0 }, "quality": { "snrDb": { "median": 54.4, "min": 28.3, "max": 81.7 }, "noiseFloorDb": -70.2, "speechLevelDb": -16.8, "bandwidthHz": 7950, "clippingShare": 0.00009, "label": "clean", "dnsmos": { "sig": 3.04, "bak": 3.42, "ovrl": 2.56 }, "method": "Frame RMS at 20 ms on the channels mixed to mono. Noise floor is the 10th percentile, speech level the 90th; SNR is their difference. Bandwidth is the highest frequency at which speech still rises 6 dB above the recording's own noise spectrum." }, "speakers": [ { "role": "Speaker 1", "id": null, "gender": "male", "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": "male", "ageBand": null } ], "speakerSummary": { "distinctIds": null, "genders": { "male": 4 }, "ageBands": null }, "conversations": [ { "index": 1, "sourceId": "conv-611461cba9", "title": "Open conversation", "durationS": 243.5, "turns": 72, "words": 537, "speakers": [ { "role": "Speaker 1", "id": null, "gender": "male", "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": "male", "ageBand": null } ], "flagged": 0, "silenceShare": 0.376, "overlapShare": 0.1, "latinShare": 0.082, "wordsPerMinute": 212.2, "specimen": { "id": "hindi-speaker-diarization--4", "windowS": 40.49, "offsetS": 0.855, "segments": 15, "bytes": 255948, "sha256": "3b404188b8eb37d38bdc194814228e710bedfb34982e5024f5f318d46ebd8a5a", "dnsmos": { "sig": 3.41, "bak": 3.26, "ovrl": 2.76 } }, "quality": { "noiseFloorDb": -67.7, "speechLevelDb": -18.3, "snrDb": 49.4, "clippingShare": 0, "bandwidthHz": 7900 }, "qualityLabel": "clean", "excerptSnrDb": 42.6 }, { "index": 2, "sourceId": "conv-e5cc82ef32", "title": "Open conversation", "durationS": 243.2, "turns": 34, "words": 644, "speakers": [ { "role": "Speaker 1", "id": null, "gender": "male", "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": "male", "ageBand": null } ], "flagged": 0, "silenceShare": 0.028, "overlapShare": 0.41, "latinShare": 0.081, "wordsPerMinute": 163.5, "specimen": null, "quality": { "noiseFloorDb": -68.5, "speechLevelDb": -27.9, "snrDb": 40.7, "clippingShare": 0, "bandwidthHz": 7900 }, "qualityLabel": "clean", "excerptSnrDb": 41.3 }, { "index": 3, "sourceId": "conv-c6122eef9c", "title": "Open conversation", "durationS": 248.2, "turns": 24, "words": 594, "speakers": [ { "role": "Speaker 1", "id": null, "gender": null, "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": null, "ageBand": null } ], "flagged": 0, "silenceShare": 0.099, "overlapShare": 0, "latinShare": 0.14, "wordsPerMinute": 159.4, "specimen": null, "quality": { "noiseFloorDb": -47, "speechLevelDb": -18.7, "snrDb": 28.3, "clippingShare": 0, "bandwidthHz": 7900 }, "qualityLabel": "some background", "excerptSnrDb": 26.8 }, { "index": 4, "sourceId": "conv-5dcf56d0af", "title": "Open conversation", "durationS": 245.5, "turns": 60, "words": 539, "speakers": [ { "role": "Speaker 1", "id": null, "gender": null, "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": null, "ageBand": null } ], "flagged": 0, "silenceShare": 0.214, "overlapShare": 0.02, "latinShare": 0.054, "wordsPerMinute": 167.6, "specimen": { "id": "hindi-speaker-diarization--3", "windowS": 43.118, "offsetS": 92.323, "segments": 14, "bytes": 274899, "sha256": "4db1f497f27499dbe9b040b0e7cbc8fb0a6642121cc2ecfbea4d7e48b335c756", "dnsmos": { "sig": 2.87, "bak": 3.68, "ovrl": 2.53 } }, "quality": { "noiseFloorDb": -97, "speechLevelDb": -15.2, "snrDb": 81.7, "clippingShare": 0, "bandwidthHz": 8000 }, "qualityLabel": "clean", "excerptSnrDb": 103.1 }, { "index": 5, "sourceId": "conv-d207e2d6f6", "title": "Open conversation", "durationS": 238.5, "turns": 68, "words": 591, "speakers": [ { "role": "Speaker 1", "id": null, "gender": null, "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": null, "ageBand": null } ], "flagged": 0, "silenceShare": 0.153, "overlapShare": 0.02, "latinShare": 0.124, "wordsPerMinute": 175.6, "specimen": { "id": "hindi-speaker-diarization--1", "windowS": 44.855, "offsetS": 31.486, "segments": 21, "bytes": 289858, "sha256": "ce52248430a77ce3dac29efedbc75ee4ae26ec76c4c40b32530ed82d80fe535a", "dnsmos": { "sig": 2.75, "bak": 3.23, "ovrl": 2.27 } }, "quality": { "noiseFloorDb": -71.8, "speechLevelDb": -9.1, "snrDb": 62.6, "clippingShare": 0.00009, "bandwidthHz": 8000 }, "qualityLabel": "clean", "excerptSnrDb": 66.6 }, { "index": 6, "sourceId": "conv-ac8dae0cca", "title": "Open conversation", "durationS": 239.7, "turns": 62, "words": 497, "speakers": [ { "role": "Speaker 1", "id": null, "gender": null, "ageBand": null }, { "role": "Speaker 2", "id": null, "gender": null, "ageBand": null } ], "flagged": 0, "silenceShare": 0.131, "overlapShare": 0.016, "latinShare": 0.056, "wordsPerMinute": 143.1, "specimen": { "id": "hindi-speaker-diarization--2", "windowS": 42.299, "offsetS": 0.103, "segments": 18, "bytes": 266878, "sha256": "56bdfbc00f0564026d2a0aae6b51ec33ec02b51b63acaff8aab400f2f27ec2e0", "dnsmos": { "sig": 3.14, "bak": 3.49, "ovrl": 2.67 } }, "quality": { "noiseFloorDb": -72.2, "speechLevelDb": -12.8, "snrDb": 59.4, "clippingShare": 0, "bandwidthHz": 8000 }, "qualityLabel": "clean", "excerptSnrDb": 59.8 } ], "specimens": [ "hindi-speaker-diarization--1", "hindi-speaker-diarization--2", "hindi-speaker-diarization--3", "hindi-speaker-diarization--4" ], "transcriptFormat": "json", "piiFlaggedSegments": 0, "withheld": false }, "gaps": [], "bundleId": "diar-hi", "relatedSlugs": [ "hindi-general-conversation", "hindi-call-centre-insurance" ], "measured": { "id": "diar-hi", "files": 1394, "conversations": 1394, "utterances": null, "turns": 432929, "speakers": 619, "speakerGender": { "female": 305, "male": 303 }, "channels": "1,394 mono conversation", "transcripts": 1394, "segments": 208826, "words": 2925696, "bandwidth": "wideband (8 kHz)", "bandwidthCounts": { "wideband (8 kHz)": 1394 }, "telephoneBandFiles": null, "snrMedianDb": 30.3, "container": "flac", "sampleRateHz": 16000, "bitDepth": 16, "pii": "Numbers of 4 or more digits, digit strings spoken as words, and email addresses are masked as [pii] in text and replaced by a tone in audio. First names are not masked.", "release": "v1.0", "updatedAt": "2026-09-29T00:57", "profile": { "conversations": 1394, "medianConversationMin": 13.38, "talkShareByRank": [ 0.715, 0.285 ], "silenceShare": 0.229, "overlapShare": 0.071, "turnsPerMin": 7.46, "wordsPerMin": 151.8, "codeMixShare": 0.168, "twoChannel": null }, "provisional": false }, "utterances": null, "url": "https://kenpathlabs.com/lokah/datasets/hindi-speaker-diarization" }