TitaNet: Neural Model for speaker representation with 1D Depth-wise separable convolutions and global context
This addresses speaker verification and diarization tasks for applications like security and communication, presenting a scalable and efficient model with incremental improvements over existing methods.
The paper tackles speaker representation extraction by proposing TitaNet, a neural network architecture using 1D depth-wise separable convolutions and global context, achieving state-of-the-art performance with an equal error rate of 0.68% on VoxCeleb1 and diarization error rates as low as 1.11% on CH109.
In this paper, we propose TitaNet, a novel neural network architecture for extracting speaker representations. We employ 1D depth-wise separable convolutions with Squeeze-and-Excitation (SE) layers with global context followed by channel attention based statistics pooling layer to map variable-length utterances to a fixed-length embedding (t-vector). TitaNet is a scalable architecture and achieves state-of-the-art performance on speaker verification task with an equal error rate (EER) of 0.68% on the VoxCeleb1 trial file and also on speaker diarization tasks with diarization error rate (DER) of 1.73% on AMI-MixHeadset, 1.99% on AMI-Lapel and 1.11% on CH109. Furthermore, we investigate various sizes of TitaNet and present a light TitaNet-S model with only 6M parameters that achieve near state-of-the-art results in diarization tasks.