Awesome ASV Anti-Spoofing

Table of contents

• Overview
• Publications
• Software
  • Framework
  • Evaluation
  • Clustering
  • Speaker embedding
  • Audio feature extraction
  • Audio data augmentation
  • Other sotware
• Datasets
  • Spoofing datasets
  • Phisical access training sets
  • Logical access training sets
  • Augmentation noise sources
• Conferences
• Leaderboards
• Other learning materials
  • Books
  • Tech blogs
  • Video tutorials
• Products
• Project
• Standards

Overview

This is a curated list of awesome ASV(Automatic Speaker Verification) Anti-Spoofing papers, libraries, datasets, and other resources.

The purpose of this repo is to organize the world’s resources for voice anti-spoofing, and make them universally accessible and useful.

To add items to this page, simply send a pull request. (contributing guide)

Publications

Special topics

Review & survey papers

• Advances in anti-spoofing: From the perspective of ASVspoof challenges, 2020
• Countermeasures to Replay Attacks: A Review, 2020
• Introduction to Voice Presentation Attack Detection and Recent Advances, 2019
• An Investigation of Deep-Learning Frameworks for Speaker Verification Anti-spoofing, 2017
• Spoofing and countermeasures for speaker verification A survey, 2015

Fast & light anti-spoofing

• Void: A fast and light voice liveness detection system, 2020
• Audio Replay Attack Detection with Deep Learning Frameworks, 2017

Anti-spoofing with phoneme

• Phoneme Specific Modelling and Scoring Techniques for Anti Spoofing System, 2019
• The SYSU System for the Interspeech 2015 Automatic Speaker Verification Spoofing and Countermeasures Challenge, 2015

Anti-spoofing with brain

• The Crux of Voice (In)Security: A Brain Study of Speaker Legitimacy Detection, 2019

Anti-spoofing with fieldprint

• The Catcher in the Field: A Fieldprint based Spoofing Detection for Text-Independent Speaker Verification, 2019
• You Can Hear But You Cannot Steal: Defending against Voice Impersonation Attacks on Smartphones, 2017

Anti-spoofing with articulatory gesture

• Hearing Your Voice is Not Enough: An Articulatory Gesture Based Liveness Detection for Voice Authentication, 2017
• VoiceLive: A Phoneme Localization based Liveness Detection for Voice Authentication on Smartphones, 2016

Anti-spoofing with Multi-task

• Adversarial Multi-Task Learning for Speaker Normalization in Replay Detection, 2020
• Multi-task learning of deep neural networks for joint automatic speaker verification and spoofing detection, 2019
• Anti-Spoofing Speaker Verification System with Multi-Feature Integration and Multi-Task Learning, 2019
• Replay spoofing detection system for automatic speaker verification using multi-task learning of noise classes, 2018

Anti-spoofing with smarthome

• Protecting Voice Controlled Systems Using Sound Source Identification Based on Acoustic Cues, 2018

Challenges

• ASVspoof 2021, 2021
• ASVspoof 2019, 2019
• ASVspoof 2017, 2017
• BTAS 2016, 2016
• ASVspoof 2015, 2015

Spoofing attack type

Self trigger attack

• Your Voice Assistant is Mine: How to Abuse Speakers to Steal Information and Control Your Phone, 2014

• A11y Attacks: Exploiting Accessibility in Operating Systems, 2014

Inaudible voice command attack

*IEMI Threats for Information Security: Remote Command Injection on Modern Smartphones, 2015

• DolphinAttack: Inaudible Voice Commands, 2017

Hidden voice command attack

• White-box attack

  • Hidden Voice Commands, 2016

  • Audio Adversarial Examples: Targeted Attacks on Speech-to-Text, 2017

  • Adversarial Attacks Against Automatic SpeechRecognition Systems via Psychoacoustic Hiding， 2018

• Black-box attack

  • CommanderSong: A Systematic Approach for Practical Adversarial Voice Recognition， 2018

  • SirenAttack: Generating Adversarial Audio for End-to-End Acoustic Systems， 2019

• Gray-box attack

  • Adversarial Music: Real World Audio AdversaryAgainst Wake-word Detection System， 2019

Voice conversion attack

• Voice conversion versus speaker verification:an overview, 2014

Speech synthesis attack

• Evaluation of Speaker Verification Security and Detection of HMM-Based Synthetic Speech, 2012

Replay attack

• A study on replay attack and anti-spoofing for text-dependent speaker verification, 2014
• A Study on Replay Attack and Anti-Spoofing for Automatic Speaker Verification, 2017

Impostor attack

• Can a Professional Imitator Fool a GMM-Based Speaker Verification System?, 2005

• I-Vectors Meet Imitators: On Vulnerability of Speaker Verification Systems Against Voice Mimicry, 2013

Other

2021

• Data Quality as Predictor of Voice Anti-Spoofing Generalization

2020

• End-to-end anti-spoofing with RawNet2
• Residual networks for resisting noise: analysis of an embeddings-based spoofing countermeasure
• Detecting Replay Attacks Using Multi-Channel Audio: A Neural Network-Based Method
• An analysis of speaker dependent models in replay detection
• Tandem Assessment of Spoofing Countermeasures and Automatic Speaker Verification: Fundamentals
• Dynamically Mitigating Data Discrepancy with Balanced Focal Loss for Replay Attack Detection
• Voice Spoofing Detection Corpus for Single and Multi-order Audio Replays
• An Ensemble Based Approach for Generalized Detection of Spoofing Attacks to Automatic Speaker Recognizers
• Defense against adversarial attacks on spoofing countermeasures of ASV
• Multiple Points Input For Convolutional Neural Networks in Replay Attack Detection

2019

• Auditory Inspired Spatial Differentiation for Replay Spoofing Attack Detection
• Attention-Based LSTM Algorithm for Audio Replay Detection in Noisy Environments
• Cross-domain replay spoofing attack detection using domain adversarial training
• Transmission Line Cochlear Model Based AM-FM Features for Replay Attack Detection
• Adversarial Attacks on Spoofing Countermeasures of automatic speaker verification
• Replay Spoofing Countermeasure Using Autoencoder and Siamese Network on ASVspoof 2019 Challenge

2018

• Independent Modelling of Long and Short Term Speech Information for Replay Detection
• Voice livness detection based on pop-noise detector with phoneme information for speaker verification
• An end-to-end spoofing countermeasure for automatic speaker verificationusing evolving recurrent neural networks
• Deep Siamese Architecture Based Replay Detection for Secure VoiceBiometric
• Use of Claimed Speaker Models for Replay Detection
• Replay Attacks Detection Using Phase and Magnitude Features with Various Frequency Resolutions
• Performance evaluation of front- and back-end techniques for ASV spoofingdetection systems based on deep features
• Modulation Dynamic Features for the Detection of Replay Attacks

2017

• Audio Replay Attack Detection Using High-Frequency Features
• Replay Attack Detection Using DNN for Channel Discrimination
• Investigating the use of Scattering Coefficients for Replay Attack Detection
• Constant Q cepstral coefficients: a spoofing countermeasure for automatic speaker verification

2016

• Anti-spoofing Methods for Automatic Speaker Verification System
• Overview of BTAS 2016 Speaker Anti-spoofing Competition
• Voice Liveness Detection for Speaker Verification based on a Tandem Single/Double-channel Pop Noise Detector
• Cross-Database Evaluation of Audio-Based Spoofing Detection Systems
• Spoofing detection from a feature representationperspective
• Spoofing Speech Detection using Temporal Convolutional Neural Network

2015

• Robust Deep Feature for Spoofing Detection - The SJTU System for ASVspoof 2015 Challenge
• A Comparison of Features for Synthetic Speech Detection

2014

• Anti-spoofing: voice databases

2013

• Vulnerability evaluation of speaker verification under voice conversionspoofing: the effect of text constraints

1999

• Vulnerability In Speaker Verification - A Study Of Technical Impostor Techniques

Software

Framework

Link	Language	Description
SpeechBrain	Python & PyTorch	SpeechBrain is an open-source and all-in-one speech toolkit based on PyTorch.
SIDEKIT	Python	SIDEKIT is an open source package allow a rapid prototyping of an end-to-end speaker recognition system.
pyAudioAnalysis	Python	Python Audio Analysis Library: Feature Extraction, Classification, Segmentation and Applications.
kaldi-asr	C++ & Bash	A toolkit for speech & speaker recognition, intended for use by researchers and professionals.
Alize LIA_SpkDet	C++	ALIZÉ is an opensource platform for speaker recognition. LIA_SpkSeg is the tools for model training,feature normalization,socre normalization,etc.
SPEAR Toolkit (based on BOB)	python	This package is part of the signal-processing and machine learning toolbox Bob.
MSRidentity Toolbox	Matlab	This toolbox contains a collection of MATLAB tools and routines that can be used for research and development in speaker recognition. PDF

Evaluation

• t-DCF: a Detection Cost Function for the Tandem Assessment of Spoofing Countermeasures and Automatic Speaker Verification [MATLAB]&[Python]
• Asvspoof 2021 Evaluation Plan

Clustering

Link	Language	Description
sklearn.cluster	Python	scikit-learn clustering algorithms.
PLDA	Python	Probabilistic Linear Discriminant Analysis & classification, written in Python.
PLDA	C++	Open-source implementation of simplified PLDA (Probabilistic Linear Discriminant Analysis).
Auto-Tuning Spectral Clustering	Python	Auto-tuning Spectral Clustering method that does not need development set or supervised tuning.

Speaker embedding

Link	Method	Language	Description
resemble-ai/Resemblyzer	d-vector	Python & PyTorch	PyTorch implementation of generalized end-to-end loss for speaker verification, which can be used for voice cloning and diarization.
Speaker_Verification	d-vector	Python & TensorFlow	Tensorflow implementation of generalized end-to-end loss for speaker verification.
PyTorch_Speaker_Verification	d-vector	Python & PyTorch	PyTorch implementation of "Generalized End-to-End Loss for Speaker Verification" by Wan, Li et al. With UIS-RNN integration.
Real-Time Voice Cloning	d-vector	Python & PyTorch	Implementation of "Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis" (SV2TTS) with a vocoder that works in real-time.
deep-speaker	d-vector	Python & Keras	Third party implementation of the Baidu paper Deep Speaker: an End-to-End Neural Speaker Embedding System.
x-vector-kaldi-tf	x-vector	Python & TensorFlow & Perl	Tensorflow implementation of x-vector topology on top of Kaldi recipe.
kaldi-ivector	i-vector	C++ & Perl	Extension to Kaldi implementing the standard i-vector hyperparameter estimation and i-vector extraction procedure.
voxceleb-ivector	i-vector	Perl	Voxceleb1 i-vector based speaker recognition system.
pytorch_xvectors	x-vector	Python & PyTorch	PyTorch implementation of Voxceleb x-vectors. Additionaly, includes meta-learning architectures for embedding training. Evaluated with speaker diarization and speaker verification.
ASVtorch	i-vector	Python & PyTorch	ASVtorch is a toolkit for automatic speaker recognition.

Audio feature extraction

Link	Language	Description
LibROSA	Python	Python library for audio and music analysis. https://librosa.github.io/
python_speech_features	Python	This library provides common speech features for ASR including MFCCs and filterbank energies. https://python-speech-features.readthedocs.io/en/latest/
pyAudioAnalysis	Python	Python Audio Analysis Library: Feature Extraction, Classification, Segmentation and Applications.

Audio data augmentation

Link	Language	Description
pyroomacoustics	Python	Pyroomacoustics is a package for audio signal processing for indoor applications. It was developed as a fast prototyping platform for beamforming algorithms in indoor scenarios. https://pyroomacoustics.readthedocs.io
gpuRIR	Python	Python library for Room Impulse Response (RIR) simulation with GPU acceleration
rir_simulator_python	Python	Room impulse response simulator using python

Other software

Link	Language	Description
Rawnet2	Python & Bash	End-to-End Neural Anti-spoofing.
ReMASC	Python	Realistic Replay Attack Corpus for Voice Controlled Systems.
Attentive-Filtering-Network	Python & Bash	University of Edinbrugh-Johns Hopkins University's system for ASVspoof 2017 Version 2.0 dataset.

Datasets

Spoofing datasets

Audio	Type	Language	Pricing	Additional information
ASVspoof 2019	PA(16.44Gb) , LA(7.116Gb)	en	Free	Evaluation Plan
ASVspoof 2017	PA-Train(200.7Mb), Dev(133.7Mb), Eval(1.065Gb)	en	Free	Evaluation Plan
SAS Corpus	LA-SS_LARGE-16k (7.591Gb), SS_LARGE-48k (7.798Gb), SS_MARY_LARGE (7.303Gb), SS_SMALL-16k (7.582Gb), SS_SMALL-16k (7.582Gb), SS_SMALL-48k (7.788Gb), VC_C1 (10.00Gb), VC_EVC (6.518Gb), VC_FESTVOX (10.04Gb), VC_FS (10.15Gb), VC_GMM (9.830Gb), VC_KPLS (9.703Gb), VC_LSP (9.616Gb), VC_TVC (6.489Gb), human (3.229Gb)	en	Free	LICENSE
ASVspoof 2015	LA-Data - Part aa (7.543Gb),Data - Part ab (7.543Gb),Data - Part ac (7.331Gb)	en	Free	LICENSE

Phisical access training sets

• ASV2019 Training set
• ASV2017 Training set

Logical access training sets

• ASV2019 Training set
• ASV2015 Training set Part aa,ab,ac

Augmentation noise sources

Name	Utterances	Pricing	Additional information
AudioSet	2M	Free	A large-scale dataset of manually annotated audio events.
MUSAN	N/A	Free	MUSAN is a corpus of music, speech, and noise recordings.

Speaker Verification training sets

Name	Utterances	Speakers	Language	Pricing	Additional information
TIMIT	6K+	630	en	$250.00	Published in 1993, the TIMIT corpus of read speech is one of the earliest speaker recognition datasets.
VCTK	43K+	109	en	Free	Most were selected from a newspaper plus the Rainbow Passage and an elicitation paragraph intended to identify the speaker's accent.
LibriSpeech	292K	2K+	en	Free	Large-scale (1000 hours) corpus of read English speech.
Multilingual LibriSpeech (MLS)	?	?	en, de, nl, es, fr, it, pt, po	Free	Multilingual LibriSpeech (MLS) dataset is a large multilingual corpus suitable for speech research. The dataset is derived from read audiobooks from LibriVox and consists of 8 languages - English, German, Dutch, Spanish, French, Italian, Portuguese, Polish.
LibriVox	180K	9K+	Multiple	Free	Free public domain audiobooks. LibriSpeech is a processed subset of LibriVox. Each original unsegmented utterance could be very long.
VoxCeleb 1&2	1M+	7K	Multiple	Free	VoxCeleb is an audio-visual dataset consisting of short clips of human speech, extracted from interview videos uploaded to YouTube.
The Spoken Wikipedia Corpora	5K	879	en, de, nl	Free	Volunteer readers reading Wikipedia articles.
CN-Celeb	130K+	1K	zh	Free	A Free Chinese Speaker Recognition Corpus Released by CSLT@Tsinghua University.
BookTubeSpeech	8K	8K	en	Free	Audio samples extracted from BookTube videos - videos where people share their opinions on books - from YouTube. The dataset can be downloaded using BookTubeSpeech-download.
DeepMine	540K	1850	fa, en	Unknown	A speech database in Persian and English designed to build and evaluate speaker verification, as well as Persian ASR systems.
NISP-Dataset	?	345	hi, kn, ml, ta, te (all Indian languages)	Free	This dataset contains speech recordings along with speaker physical parameters (height, weight, ... ) as well as regional information and linguistic information.

Conferences

Conference/Workshop	Frequency	Page Limit	Organization	Blind Review
ICASSP	Annual	4 + 1 (ref)	IEEE	No
InterSpeech	Annual	4 + 1 (ref)	ISCA	No
APSIPA	Annual	4 + 1 (ref)	IEEE	Yes
Odyssey	Biennial	8 + 2 (ref)	ISCA	No
SLT	Biennial	6 + 2 (ref)	IEEE	Yes
ASRU	Biennial	6 + 2 (ref)	IEEE	Yes
WASPAA	Biennial	4 + 1 (ref)	IEEE	No

Other learning materials

Books

• Handbook of Biometric Anti-Spoofing

Tech blogs

• Can You Fool Voice Biometrics? by Lior Artzi

• ID R&D and Synaptics First to Deploy Voice Biometrics on NPU for Smart Home Applications by Vineet Ganju

Video tutorials

Products

Company	Product
Pindrop	Deep Voice Engine
ID R&D	IDLive™ Voice
VoiceAI	Voiceprint recognition API
Kriston	Voiceprint API,SDK

Project

• OCTAVE

Standards

• Information security technology — Security requirements of voiceprint recognition data(Exposure draft), 2021

• Technical specifications for voiceprint recognition based security application for mobile finance (JR / t0164-2018), 2018