Towards generalisable and calibrated synthetic speech detection with self-supervised representations

no code implementations • 11 Sep 2023 • Dan Oneata, Adriana Stan, Octavian Pascu, Elisabeta Oneata, Horia Cucu

Generalisation -- the ability of a model to perform well on unseen data -- is crucial for building reliable deep fake detectors.

Synthetic Speech Detection

Adaptation of Whisper models to child speech recognition

1 code implementation • 24 Jul 2023 • Rishabh Jain, Andrei Barcovschi, Mariam Yiwere, Peter Corcoran, Horia Cucu

We demonstrate that finetuning Whisper on child speech yields significant improvements in ASR performance on child speech, compared to non finetuned Whisper models.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

Adaptive Planning Search Algorithm for Analog Circuit Verification

no code implementations • 23 Jun 2023 • Cristian Manolache, Cristina Andronache, Alexandru Caranica, Horia Cucu, Andi Buzo, Cristian Diaconu, Georg Pelz

Thus, we show that the proposed approach is able to provide OCCs closer to the specifications for all circuits and identify a failure (specification violation) for one of the responses of a real circuit.

FlexLip: A Controllable Text-to-Lip System

no code implementations • 7 Jun 2022 • Dan Oneata, Beata Lorincz, Adriana Stan, Horia Cucu

This modularity enables the easy replacement of each of its components, while also ensuring the fast adaptation to new speaker identities by disentangling or projecting the input features.

Audio Generation Text-to-Video Generation +1

Automated Circuit Sizing with Multi-objective Optimization based on Differential Evolution and Bayesian Inference

no code implementations • 6 Jun 2022 • Catalin Visan, Octavian Pascu, Marius Stanescu, Elena-Diana Sandru, Cristian Diaconu, Andi Buzo, Georg Pelz, Horia Cucu

The proposed method is able to perform sizing for complex circuits with a large number of design variables and many conflicting objectives to be optimized.

Bayesian Inference Bayesian Optimization +2

A Wav2vec2-Based Experimental Study on Self-Supervised Learning Methods to Improve Child Speech Recognition

no code implementations • 6 Apr 2022 • Rishabh Jain, Andrei Barcovschi, Mariam Yiwere, Dan Bigioi, Peter Corcoran, Horia Cucu

Our models outperformed the wav2vec2 BASE 960 on child speech which is considered a state-of-the-art ASR model on adult speech by just using 10 hours of child speech data in finetuning.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +2

A Text-to-Speech Pipeline, Evaluation Methodology, and Initial Fine-Tuning Results for Child Speech Synthesis

no code implementations • 22 Mar 2022 • Rishabh Jain, Mariam Yiwere, Dan Bigioi, Peter Corcoran, Horia Cucu

Speech synthesis has come a long way as current text-to-speech (TTS) models can now generate natural human-sounding speech.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +3

Speaker disentanglement in video-to-speech conversion

1 code implementation • 20 May 2021 • Dan Oneata, Adriana Stan, Horia Cucu

The task of video-to-speech aims to translate silent video of lip movement to its corresponding audio signal.

Disentanglement Speech Synthesis

An evaluation of word-level confidence estimation for end-to-end automatic speech recognition

no code implementations • 14 Jan 2021 • Dan Oneata, Alexandru Caranica, Adriana Stan, Horia Cucu

In this paper we investigate confidence estimation for end-to-end automatic speech recognition (ASR).

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

RSC: A Romanian Read Speech Corpus for Automatic Speech Recognition

no code implementations • LREC 2020 • Alex Georgescu, ru-Lucian, Horia Cucu, Andi Buzo, Corneliu Burileanu

Although many efforts have been made in the last decade to enhance the speech and language resources for Romanian, this language is still considered under-resourced.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

The Quo Vadis submission at Traffic4cast 2019

1 code implementation • 27 Oct 2019 • Dan Oneata, Cosmin George Alexandru, Marius Stanescu, Octavian Pascu, Alexandru Magan, Adrian Postelnicu, Horia Cucu

We describe the submission of the Quo Vadis team to the Traffic4cast competition, which was organized as part of the NeurIPS 2019 series of challenges.

regression

Kite: Automatic speech recognition for unmanned aerial vehicles

no code implementations • 2 Jul 2019 • Dan Oneata, Horia Cucu

This paper addresses the problem of building a speech recognition system attuned to the control of unmanned aerial vehicles (UAVs).

Automatic Speech Recognition Automatic Speech Recognition (ASR) +2