RawTFNet: A Lightweight CNN Architecture for Speech Anti-Spoofing
Authors:
- 22 October 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1997-2001
Abstract
Automatic speaker verification (ASV) systems are often affected by spoofing attacks. Recent transformer-based models have improved anti-spoofing performance by learning strong feature representations. However, these models usually need high computing power. To address this, we introduce RawTFNet, a lightweight CNN model designed for audio signals. The RawTFNet separates feature processing along time and frequency dimensions, which helps to capture the fine-grained details of synthetic speech. We tested RawTFNet on the ASVspoof 2021 LA and DF evaluation datasets. The results show that RawTFNet reaches comparable performance to that of the state-of-the-art models, while also using fewer computing resources. The code and models will be made publicly available.Keywords
This publication has 27 references indexed in Scilit:
- Refining the evaluation of speech synthesis: A summary of the Blizzard Challenge 2023Computer Speech & Language, 2024
- TF-SepNet: An Efficient 1D Kernel Design in Cnns for Low-Complexity Acoustic Scene Classification2024
- Leveraging Positional-Related Local-Global Dependency for Synthetic Speech Detection2023
- AASIST: Audio Anti-Spoofing Using Integrated Spectro-Temporal Graph Attention Networks2022
- Rawboost: A Raw Data Boosting and Augmentation Method Applied to Automatic Speaker Verification Anti-Spoofing2022
- End-to-End anti-spoofing with RawNet22021
- Squeeze-and-Excitation Networks2018
- Constant Q cepstral coefficients: A spoofing countermeasure for automatic speaker verificationComputer Speech & Language, 2017
- ASVspoof: The Automatic Speaker Verification Spoofing and Countermeasures ChallengeIEEE Journal of Selected Topics in Signal Processing, 2017
- Spoofing and countermeasures for speaker verification: A surveySpeech Communication, 2015