Search for: All records

AI-synthesized voice technology has the potential to create realistic human voices for beneficial applications, but it can also be misused for malicious purposes. While existing AI-synthesized voice detection models excel in intra-domain evaluation, they face challenges in generalizing across different domains, potentially becoming obsolete as new voice generators emerge. Current solutions use diverse data and advanced machine learning techniques (e.g., domain-invariant representation, self-supervised learning), but are limited by predefined vocoders and sensitivity to factors like background noise and speaker identity. In this work, we introduce an innovative disentanglement framework aimed at extracting domain-agnostic artifact features related to vocoders. Utilizing these features, we enhance model learning in a flat loss landscape, enabling escape from suboptimal solutions and improving generalization. Extensive experiments on benchmarks show our approach outperforms state-of-the-art methods, achieving up to 5.12% improvement in the equal error rate metric in intra-domain and 7.59% in cross-domain evaluations. 

We report visible-enhanced stimulated Raman scattering (VIS-SRS) microscopy by direct frequency-doubling of picosecond pulsed lasers. Compared to previous reports from femtosecond sources, our setup is greatly simplified by removing the need to perform spectral focusing and allows for quick wavelength tuning for sparse multi-channel imaging across a large spectral range. We report a signal enhancement of 3.2-fold for the C–H bond from decane, consistent with theoretical analysis, and up to 483-fold with resonance enhancement. We showcase a multi-channel imaging application in HeLa cells and demonstrate a spatial resolution down to 47 nm from 5xFAD mouse brain tissues coupled with Raman-tailored sample-expansion. Finally, we discuss the future applications and limitations of our picosecond-doubled VIS-SRS.