Ensemble learning model for deepfake audio detection using multi-feature extraction approach
Parna Chaudhury, L Shivani Narayan, S Mridhula, Ravi Tiwari · 2026
Read Ensemble learning model for deepfake audio detection using multi-feature extraction approach by Parna Chaudhury, L Shivani Narayan, S Mridhula, Ravi Tiwari (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13572470.txt.
Description
Parna Chaudhury, with three colleagues, presents a 2026 system for spotting fake audio, where a cloned voice says things a person never said. Such clips fuel fraud and false claims, and they are hard to catch by ear. The system reads a clip in three ways at once to catch traces that one view might miss.
One branch reads sound spectra over time, one reads cepstral features that track the voice's shape, and one reads plain statistics of the signal. Three network types handle the branches, and a final model stacks their calls into one verdict. On the Fake-or-Real data set, the stacked system beat each single model. The authors also host the tool online for live checks. They frame fused features as the key to fewer missed fakes.
Catalog record
| CREATOR | Parna Chaudhury, L Shivani Narayan, S Mridhula, Ravi Tiwari |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48373 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |