VoxConverse

Dataset Card for the Voxconverse dataset VoxConverse is an audio-visual diarisation dataset consisting of multispeaker clips of human speech, extracted from YouTube videos. Updates and additional information about the dataset can be found on the dataset website. Note: This dataset has been preprocessed using diarizers.

Papers using VoxConverse (9)

Whisper Speaker Identification: Leveraging Pre-Trained Multilingual Transformers for Robust Speaker Embeddings2025 · 1 citesFast and Robust On-Device Speaker Diarization: Relative Minimum Cluster Size for Stride-Accelerated Pipelines2026Supervised Hierarchical Clustering using Graph Neural Networks for Speaker Diarization2023 · 12 citesTarget-Speaker Voice Activity Detection via Sequence-to-Sequence Prediction2022 · 2 citesTarget Speaker Voice Activity Detection with Transformers and Its Integration with End-to-End Neural Diarization2022 · 1 citesMulti-Input Multi-Output Target-Speaker Voice Activity Detection For Unified, Flexible, and Robust Audio-Visual Speaker Diarization2024 · 1 citesProfile-Error-Tolerant Target-Speaker Voice Activity Detection2023End-to-End Supervised Hierarchical Graph Clustering for Speaker Diarization2024Speaker Embeddings With Weakly Supervised Voice Activity Detection For Efficient Speaker Diarization2024