Weakly Supervised Training Of Speaker Identification Models

Abstract

We propose an approach for training speaker identification models in a weakly supervised manner. We concentrate on the setting where the training data consists of a set of audio recordings and the speaker annotation is provided only at the recording level. The method uses speaker diarization to find unique speakers in each recording, and i-vectors to project the speech of each speaker to a fixed-dimensional vector. A neural network is then trained to map i-vectors to speakers, using a special objective function that allows to optimize the model using recording-level speaker labels. We report experiments on two different real-world datasets. On the VoxCeleb dataset, the method provides 94.6% accuracy on a closed set speaker identification task, surpassing the baseline performance by a large margin. On an Estonian broadcast news dataset, the method provides 66% time-weighted speaker identification recall at 93% precision.

Stats

• citations5
• S2 citations—
• github stars0
• HF likes0
• heat score5.84
• arxiv keykaru2018weakly

Related papers

• Training Speaker Embedding Extractors Using Multi-speaker Audio With Unknown Speaker Boundaries (2022)3.58
• T-vectors: Weakly Supervised Speaker Identification Using Hierarchical Transformer Model (2020)0.00
• Weakly Supervised Training Of Hierarchical Attention Networks For Speaker Identification (2020)3.58
• Training Speaker Recognition Systems With Limited Data (2022)8.13
• Robust Speaker Recognition Using Unsupervised Adversarial Invariance (2019)9.76
• Speaker Embeddings With Weakly Supervised Voice Activity Detection For Efficient Speaker Diarization (2024)2.26
• State-of-the-art Embeddings With Video-free Segmentation Of The Source Voxceleb Data (2024)0.00
• Channel Adversarial Training For Speaker Verification And Diarization (2019)7.50