AI Chat AI Image Generator AI Video Text to Speech

Training strategy for a lightweight countermeasure model for automatic speaker verification

03/31/2022

∙

by Yen-Lun Liao, et al.

∙

∙

The countermeasure (CM) model is developed to protect Automatic Speaker Verification (ASV) systems from spoof attacks and prevent resulting personal information leakage. Based on practicality and security considerations, the CM model is usually deployed on edge devices, which have more limited computing resources and storage space than cloud-based systems. This work proposes training strategies for a lightweight CM model for ASV, using generalized end-to-end (GE2E) pre-training and adversarial fine-tuning to improve performance, and applying knowledge distillation (KD) to reduce the size of the CM model. In the evaluation phase of the ASVspoof 2021 Logical Access task, the lightweight ResNetSE model reaches min t-DCF 0.2695 and EER 3.54 the teacher model, the lightweight student model only uses 22.5 and 21.1

Yen-Lun Liao
1 publication
Xuanjun Chen
4 publications
Chung-Che Wang
3 publications
Jyh-Shing Roger Jang
9 publications

page 1

page 2

page 3

page 4

research

∙ 05/27/2023

One-Step Knowledge Distillation and Fine-Tuning in Using Large Pre-Trained Self-Supervised Learning Models for Speaker Verification

The application of speech self-supervised learning (SSL) models has achi...

0 Jungwoo Heo, et al. ∙

research

∙ 04/27/2020

LightPAFF: A Two-Stage Distillation Framework for Pre-training and Fine-tuning

While pre-training and fine-tuning, e.g., BERT <cit.>, GPT-2 <cit.>, hav...

0 Kaitao Song, et al. ∙

research

∙ 08/17/2023

Learning Lightweight Object Detectors via Multi-Teacher Progressive Distillation

Resource-constrained perception systems such as edge computing and visio...

0 Shengcao Cao, et al. ∙

research

∙ 10/09/2021

Towards Lightweight Applications: Asymmetric Enroll-Verify Structure for Speaker Verification

With the development of deep learning, automatic speaker verification ha...

0 Qingjian Lin, et al. ∙

research

∙ 10/14/2022

Lightweight Alpha Matting Network Using Distillation-Based Channel Pruning

Recently, alpha matting has received a lot of attention because of its u...

0 Donggeun Yoon, et al. ∙

research

∙ 09/21/2020

Open-set Short Utterance Forensic Speaker Verification using Teacher-Student Network with Explicit Inductive Bias

In forensic applications, it is very common that only small naturalistic...

0 Mufan Sang, et al. ∙

research

∙ 07/03/2021

Pool of Experts: Realtime Querying Specialized Knowledge in Massive Neural Networks

In spite of the great success of deep learning technologies, training an...

0 Hakbin Kim, et al. ∙