Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Boris Ginsburg

Boris Ginsburg

VANI: Very-lightweight Accent-controllable TTS for Native and Non-native speakers with Identity Preservation

Mar 14, 2023
Rohan Badlani, Akshit Arora, Subhankar Ghosh, Rafael Valle, Kevin J. Shih, João Felipe Santos, Boris Ginsburg, Bryan Catanzaro

Figure 1 for VANI: Very-lightweight Accent-controllable TTS for Native and Non-native speakers with Identity Preservation

* Presentation accepted at ICASSP 2023

Access Paper or Ask Questions

ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations

Feb 16, 2023
Shehzeen Hussain, Paarth Neekhara, Jocelyn Huang, Jason Li, Boris Ginsburg

Figure 1 for ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations

Figure 2 for ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations

Figure 3 for ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations

Figure 4 for ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations

* Published as a conference paper at ICASSP 2023

Access Paper or Ask Questions

Fast Entropy-Based Methods of Word-Level Confidence Estimation for End-To-End Automatic Speech Recognition

Dec 16, 2022
Aleksandr Laptev, Boris Ginsburg

Figure 1 for Fast Entropy-Based Methods of Word-Level Confidence Estimation for End-To-End Automatic Speech Recognition

Figure 2 for Fast Entropy-Based Methods of Word-Level Confidence Estimation for End-To-End Automatic Speech Recognition

Figure 3 for Fast Entropy-Based Methods of Word-Level Confidence Estimation for End-To-End Automatic Speech Recognition

Figure 4 for Fast Entropy-Based Methods of Word-Level Confidence Estimation for End-To-End Automatic Speech Recognition

* To appear in Proc. SLT 2022, Jan 09-12, 2023, Doha, Qatar. 8 pages, 4 figures, 4 tables

Access Paper or Ask Questions

Accidental Learners: Spoken Language Identification in Multilingual Self-Supervised Models

Nov 09, 2022
Travis M. Bartley, Fei Jia, Krishna C. Puvvada, Samuel Kriman, Boris Ginsburg

Figure 1 for Accidental Learners: Spoken Language Identification in Multilingual Self-Supervised Models

Figure 2 for Accidental Learners: Spoken Language Identification in Multilingual Self-Supervised Models

Figure 3 for Accidental Learners: Spoken Language Identification in Multilingual Self-Supervised Models

Figure 4 for Accidental Learners: Spoken Language Identification in Multilingual Self-Supervised Models

* Submitted to ICASSP 2023

Access Paper or Ask Questions

Multi-blank Transducers for Speech Recognition

Nov 04, 2022
Hainan Xu, Fei Jia, Somshubra Majumdar, Shinji Watanabe, Boris Ginsburg

Figure 1 for Multi-blank Transducers for Speech Recognition

Figure 2 for Multi-blank Transducers for Speech Recognition

Figure 3 for Multi-blank Transducers for Speech Recognition

Figure 4 for Multi-blank Transducers for Speech Recognition

* Submitted to ICASSP 2023

Access Paper or Ask Questions

Adapter-Based Extension of Multi-Speaker Text-to-Speech Model for New Speakers

Nov 01, 2022
Cheng-Ping Hsieh, Subhankar Ghosh, Boris Ginsburg

Figure 1 for Adapter-Based Extension of Multi-Speaker Text-to-Speech Model for New Speakers

Figure 2 for Adapter-Based Extension of Multi-Speaker Text-to-Speech Model for New Speakers

Figure 3 for Adapter-Based Extension of Multi-Speaker Text-to-Speech Model for New Speakers

Figure 4 for Adapter-Based Extension of Multi-Speaker Text-to-Speech Model for New Speakers

* Submitted to ICASSP 2023

Access Paper or Ask Questions

AmberNet: A Compact End-to-End Model for Spoken Language Identification

Oct 27, 2022
Fei Jia, Nithin Rao Koluguri, Jagadeesh Balam, Boris Ginsburg

Figure 1 for AmberNet: A Compact End-to-End Model for Spoken Language Identification

Figure 2 for AmberNet: A Compact End-to-End Model for Spoken Language Identification

Figure 3 for AmberNet: A Compact End-to-End Model for Spoken Language Identification

Figure 4 for AmberNet: A Compact End-to-End Model for Spoken Language Identification

* Submitted to ICASSP 2023

Access Paper or Ask Questions

Damage Control During Domain Adaptation for Transducer Based Automatic Speech Recognition

Oct 06, 2022
Somshubra Majumdar, Shantanu Acharya, Vitaly Lavrukhin, Boris Ginsburg

Figure 1 for Damage Control During Domain Adaptation for Transducer Based Automatic Speech Recognition

Figure 2 for Damage Control During Domain Adaptation for Transducer Based Automatic Speech Recognition

Figure 3 for Damage Control During Domain Adaptation for Transducer Based Automatic Speech Recognition

* To appear in Proc. SLT 2022, Jan 09-12, 2023, Doha, Qatar

Access Paper or Ask Questions