Stereo multi-speaker Hindi dialogue recordings with L/R speaker separation and emotion annotations
Languages
Hindi
Quality Check
100% Verified
Natural Hindi conversational recordings featuring spontaneous two-speaker dialogues between native Hindi speakers. Stereo recordings with dedicated left/right audio channels per speaker for clean speaker separation—recorded via LiveKit with isolated per-participant tracks. Covers everyday conversations, personal discussions, and topical debates with speakers from across the Hindi belt including Delhi, Mumbai, Uttar Pradesh, and Bihar dialect regions. Features speaker diarization, word-level timestamps, and per-utterance emotion labels. Ideal for training multi-speaker Hindi ASR, speaker extraction, diarization, and conversational AI models.
Stereo WAV files with L/R speaker separation (48kHz, 16-bit), JSON transcripts with speaker diarization labels, Word-level timestamps per speaker, Per-utterance emotion labels with confidence scores, Speaker metadata
48kHz sample rate, 16-bit depth, 1536 kbps, stereo with L/R channel separation
JSON with speaker labels, word timestamps, emotion annotations