Audit of 39 deepfake speech datasets reveals ethical and technical limits

Arxiv CS.CV (computer vision)

Paper audits 39 deepfake speech datasets, finding that claims about detector robustness and fairness are limited by dataset quality. Authors call for more representative and ethically sourced data.

DeepfakesSpeech

See 14 more sources

SpAArSIST: Sparsified AASIST for Efficient and Reliable Anti-Spoofing3 days agoAnton Firc, Vojt\v{e}ch Stan\v{e}k, Zbyn\v{e}k Li\v{c}ka, Kamil Malinka, Martin Pere\v{s}\'ini

The CIFAR Synthetic Evidence Corpus for Detecting AI-Generated Evidence5 days agoKelly McConvey, Jalehsadat Mahdavimoghaddam, Nima Jamali, Maksym Taranukhin, Sajad Ebrahimi, Wentao Zhang, Yuntian Deng, Karen Eltis, Maura R. Grossman, Vered Shwartz, Ebrahim Bagheri

Exploring the Scale and Diversity of Speech Anti-spoofing Datasets: Experiments and Analysis5 days agoZhuolin Yi, Jun Xue, Yanzhen Ren, Yihuan Huang, Yi Chai, Daixian Li, Guanxiang Feng, Jiajun Liu

Dissect and Prune: Enhancing Robustness in AI-Generated Image Detection4 days agoDahye Kim, Jaehyun Choi, Hyun Seok Seong, Seongho Kim, Donghun Lee, Sungwon Yi, Jang-Ho Choi

Assessing the Energy and Carbon Emissions of Neural Speaker Verification Model in Training and Inference5 days agoHugo Leguillier, Driss Matrouf, Guillaume Lechien, Mickael Rouvier

A Comparison of SSL-Based Feature Extractors and Back-End Classifiers for Spoofing Detection: A Multi-Corpus Training and Cross-Linguistic Analysis5 days agoAnh-Tuan Dao, Driss Matrouf, Mickael Rouvier, Nicholas Evans

Speaker-Invariant Representation Learning for Spoofing Detection via Gradient Reversal and A Variational Information Bottleneck5 days agoAnh-Tuan Dao, Driss Matrouf, Mickael Rouvier, Nicholas Evans

Overview of ESDD2: Environment-Aware Speech and Sound Deepfake Detection Challenge4 days agoXueping Zhang, Han Yin, Yang Xiao, Lin Zhang, Ting Dang, Rohan Kumar Das, Ming Li

RAT: Reference-Augmented Training for ASV Anti-Spoofing4 days agoVojt\v{e}ch Stan\v{e}k, Anton Firc, Jakub Re\v{s}, Kamil Malinka

Ethical and Technical Limits of Deepfake Speech Datasets4 days agoVojt\v{e}ch Stan\v{e}k, Eva Trnovsk\'a, Kamil Malinka, Anton Firc

Toward Calibrated, Fair, and accurate Deepfake Detection4 days agoRyan Brown, Chris Russell

MultiAPI Spoof: A Multi-API Dataset and Local-Attention Network for Speech Anti-spoofing Detection4 days agoXueping Zhang, Zhenshan Zhang, Yechen Wang, Linxi Li, Liwei Jin, Ming Li

The Hidden Cost of Pairwise Verification in Synthetic Speech Source Tracing3 days agoAnton Firc, Zbyn\v{e}k Li\v{c}ka, Vojt\v{e}ch Stan\v{e}k, Kamil Malinka

On the Study of Biometric Spoofing Detection using Deep Learning3 days agoKumar Kartikey, Nikos Komninos

Arxiv CS.CV (computer vision)·Omid AhmadiehNima Karimian

4 days ago