NatiQ: An End-to-end Text-to-Speech System for Arabic

no code implementations • 15 Jun 2022 • Ahmed Abdelali, Nadir Durrani, Cenk Demiroglu, Fahim Dalvi, Hamdy Mubarak, Kareem Darwish

We concatenated Tacotron1 with the WaveRNN vocoder, Tacotron2 with the WaveGlow vocoder and ESPnet transformer with the parallel wavegan vocoder to synthesize waveforms from the spectrograms.

Investigation of Synthetic Speech Detection Using Frame- and Segment-Specific Importance Weighting

no code implementations • 10 Oct 2016 • Ali Khodabakhsh, Cenk Demiroglu

Thus, the type of distortions that were caused by the unknown systems were different and could not be captured better with the proposed SSD compared to the baseline SSD.

Speaker Verification Synthetic Speech Detection

Incorporation of Speech Duration Information in Score Fusion of Speaker Recognition Systems

no code implementations • 7 Aug 2016 • Ali Khodabakhsh, Seyyed Saeed Sarfjoo, Umut Uludag, Osman Soyyigit, Cenk Demiroglu

Here, we investigate the effect of speech duration on the performance of three state-of-the-art speaker recognition systems.

Speaker Recognition Speaker Verification