Joint Fullband-Subband Modeling for High-Resolution SingFake Detection
This addresses the urgent need for better detection of unauthorized singing voice imitations, which is an incremental improvement over existing methods.
The study tackled the problem of detecting singing voice deepfakes by analyzing high-resolution audio at 44.1 kHz, revealing that high-frequency subbands provide essential cues, and their joint fullband-subband framework significantly outperformed 16 kHz-sampled models.
Rapid advances in singing voice synthesis have increased unauthorized imitation risks, creating an urgent need for better Singing Voice Deepfake (SingFake) Detection, also known as SVDD. Unlike speech, singing contains complex pitch, wide dynamic range, and timbral variations. Conventional 16 kHz-sampled detectors prove inadequate, as they discard vital high-frequency information. This study presents the first systematic analysis of high-resolution (44.1 kHz sampling rate) audio for SVDD. We propose a joint fullband-subband modeling framework: the fullband captures global context, while subband-specific experts isolate fine-grained synthesis artifacts unevenly distributed across the spectrum. Experiments on the WildSVDD dataset demonstrate that high-frequency subbands provide essential complementary cues. Our framework significantly outperforms 16 kHz-sampled models, proving that high-resolution audio and strategic subband integration are critical for robust in-the-wild detection.