Ehsan Variani

Person

Baltimore, MD, US

People

Patents Grantslast 30 patents

Information Patent Grant
Multilingual re-scoring models for automatic speech recognition
- Patent number 12,080,283
- Issue date Sep 3, 2024
- Google LLC
- Neeraj Gaur
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Efficient streaming non-recurrent on-device end-to-end model
- Patent number 12,051,404
- Issue date Jul 30, 2024
- Google LLC
- Tara Sainath
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Exponential modeling with deep learning features
- Patent number 11,922,322
- Issue date Mar 5, 2024
- Google LLC
- Mitchel Weintraub
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Grant
Enhanced multi-channel acoustic models
- Patent number 11,783,849
- Issue date Oct 10, 2023
- Google LLC
- Ehsan Variani
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Efficient streaming non-recurrent on-device end-to-end model
- Patent number 11,715,458
- Issue date Aug 1, 2023
- Google LLC
- Tara Sainath
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Exponential modeling with deep learning features
- Patent number 11,568,260
- Issue date Jan 31, 2023
- Google LLC
- Mitchel Weintraub
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Grant
Complex linear projection for acoustic modeling
- Patent number 11,393,457
- Issue date Jul 19, 2022
- Google LLC
- Samuel Bengio
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Multichannel speech recognition using neural networks
- Patent number 11,062,725
- Issue date Jul 13, 2021
- Google LLC
- Ehsan Variani
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Linear transformation for speech recognition modeling
- Patent number 10,714,078
- Issue date Jul 14, 2020
- Google LLC
- Samuel Bengio
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Enhanced multi-channel acoustic models
- Patent number 10,224,058
- Issue date Mar 5, 2019
- Google LLC
- Ehsan Variani
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Complex linear projection for acoustic modeling
- Patent number 10,140,980
- Issue date Nov 27, 2018
- Google LCC
- Samuel Bengio
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Speaker verification using neural networks
- Patent number 9,401,148
- Issue date Jul 26, 2016
- Google Inc.
- Xin Lei
- G10 - MUSICAL INSTRUMENTS ACOUSTICS

Patents Applicationslast 30 patents

Information Patent Application
EFFICIENT STREAMING NON-RECURRENT ON-DEVICE END-TO-END MODEL
- Publication number 20240371363
- Publication date Nov 7, 2024
- Google LLC
- Tara Sainath
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
Multilingual Re-Scoring Models for Automatic Speech Recognition
- Publication number 20240203409
- Publication date Jun 20, 2024
- Google LLC
- Neeraj Gaur
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
EFFICIENT STREAMING NON-RECURRENT ON-DEVICE END-TO-END MODEL
- Publication number 20230343328
- Publication date Oct 26, 2023
- Google LLC
- Tara Sainath
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
Rare Word Recognition with LM-aware MWER Training
- Publication number 20230298570
- Publication date Sep 21, 2023
- Google LLC
- Weiran Wang
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
Exponential Modeling with Deep Learning Features
- Publication number 20230186096
- Publication date Jun 15, 2023
- Google LLC
- Mitchel Weintraub
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Application
Multilingual Re-Scoring Models for Automatic Speech Recognition
- Publication number 20220310081
- Publication date Sep 29, 2022
- Google LLC
- Neeraj Gaur
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
Efficient Streaming Non-Recurrent On-Device End-to-End Model
- Publication number 20220310062
- Publication date Sep 29, 2022
- Google LLC
- Tara Sainath
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
Cascaded Encoders for Simplified Streaming and Non-Streaming ASR
- Publication number 20220122622
- Publication date Apr 21, 2022
- Google LLC
- Arun Narayanan
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Application
ENHANCED MULTI-CHANNEL ACOUSTIC MODELS
- Publication number 20210295859
- Publication date Sep 23, 2021
- Google LLC
- Ehsan Variani
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
COMPLEX LINEAR PROJECTION FOR ACOUSTIC MODELING
- Publication number 20200286468
- Publication date Sep 10, 2020
- Google LLC
- Samuel Bengio
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
Exponential Modeling with Deep Learning Features
- Publication number 20200134466
- Publication date Apr 30, 2020
- Google LLC
- Mitchel Weintraub
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Application
ENHANCED MULTI-CHANNEL ACOUSTIC MODELS
- Publication number 20190259409
- Publication date Aug 22, 2019
- Google LLC
- Ehsan Variani
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
COMPLEX LINEAR PROJECTION FOR ACOUSTIC MODELING
- Publication number 20190115013
- Publication date Apr 18, 2019
- Google LLC
- Samuel Bengio
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
COMPLEX LINEAR PROJECTION FOR ACOUSTIC MODELING
- Publication number 20180174575
- Publication date Jun 21, 2018
- Google LLC
- Samuel Bengio
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
ENHANCED MULTI-CHANNEL ACOUSTIC MODELS
- Publication number 20180068675
- Publication date Mar 8, 2018
- Google Inc.
- Ehsan Variani
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
SPEAKER VERIFICATION USING NEURAL NETWORKS
- Publication number 20150127336
- Publication date May 7, 2015
- Google Inc.
- Xin Lei
- G10 - MUSICAL INSTRUMENTS ACOUSTICS

Ehsan Variani

Person

Patents Grantslast 30 patents

Multilingual re-scoring models for automatic speech recognition

Efficient streaming non-recurrent on-device end-to-end model

Exponential modeling with deep learning features

Enhanced multi-channel acoustic models

Efficient streaming non-recurrent on-device end-to-end model

Exponential modeling with deep learning features

Complex linear projection for acoustic modeling

Multichannel speech recognition using neural networks

Linear transformation for speech recognition modeling

Enhanced multi-channel acoustic models

Complex linear projection for acoustic modeling

Speaker verification using neural networks

Patents Applicationslast 30 patents

EFFICIENT STREAMING NON-RECURRENT ON-DEVICE END-TO-END MODEL

Multilingual Re-Scoring Models for Automatic Speech Recognition

EFFICIENT STREAMING NON-RECURRENT ON-DEVICE END-TO-END MODEL

Rare Word Recognition with LM-aware MWER Training

Exponential Modeling with Deep Learning Features

Multilingual Re-Scoring Models for Automatic Speech Recognition

Efficient Streaming Non-Recurrent On-Device End-to-End Model

Cascaded Encoders for Simplified Streaming and Non-Streaming ASR

ENHANCED MULTI-CHANNEL ACOUSTIC MODELS

COMPLEX LINEAR PROJECTION FOR ACOUSTIC MODELING

Exponential Modeling with Deep Learning Features

ENHANCED MULTI-CHANNEL ACOUSTIC MODELS

COMPLEX LINEAR PROJECTION FOR ACOUSTIC MODELING

COMPLEX LINEAR PROJECTION FOR ACOUSTIC MODELING

ENHANCED MULTI-CHANNEL ACOUSTIC MODELS

SPEAKER VERIFICATION USING NEURAL NETWORKS