Yashesh GAUR

Person

Redmond, WA, US

People

Patents Grantslast 30 patents

Information Patent Grant
Dynamic gradient aggregation for training neural networks
- Patent number 12,136,034
- Issue date Nov 5, 2024
- Microsoft Technology Licensing, LLC
- Dimitrios B. Dimitriadis
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Grant
Training and using a transcript generation model on a multi-speaker...
- Patent number 11,984,127
- Issue date May 14, 2024
- Microsoft Technology Licensing, LLC
- Naoyuki Kanda
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Hypothesis stitcher for speech recognition of long-form audio
- Patent number 11,935,542
- Issue date Mar 19, 2024
- Microsoft Technology Licensing, LLC
- Naoyuki Kanda
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Speaker adaptation for attention-based encoder-decoder
- Patent number 11,915,686
- Issue date Feb 27, 2024
- Microsoft Technology Licensing, LLC
- Zhong Meng
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Hypothesis stitcher for speech recognition of long-form audio
- Patent number 11,574,639
- Issue date Feb 7, 2023
- Microsoft Technology Licensing, LLC
- Naoyuki Kanda
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Sequence-to-sequence speech recognition with latency threshold
- Patent number 11,562,745
- Issue date Jan 24, 2023
- Microsoft Technology Licensing, LLC
- Yashesh Gaur
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Grant
Internal language model for E2E models
- Patent number 11,527,238
- Issue date Dec 13, 2022
- Microsoft Technology Licensing, LLC
- Zhong Meng
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Grant
Speaker adaptation for attention-based encoder-decoder
- Patent number 11,232,782
- Issue date Jan 25, 2022
- Microsoft Technology Licensing, LLC
- Zhong Meng
- G10 - MUSICAL INSTRUMENTS ACOUSTICS

Patents Applicationslast 30 patents

Information Patent Application
TRAINING AND USING A TRANSCRIPT GENERATION MODEL ON A MULTI-SPEAKER...
- Publication number 20240257815
- Publication date Aug 1, 2024
- Microsoft Technology Licensing, LLC
- Naoyuki KANDA
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
HYPOTHESIS STITCHER FOR SPEECH RECOGNITION OF LONG-FORM AUDIO
- Publication number 20240185859
- Publication date Jun 6, 2024
- Microsoft Technology Licensing, LLC
- Naoyuki KANDA
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
ON-DEVICE STREAMING INVERSE TEXT NORMALIZATION (ITN)
- Publication number 20230289536
- Publication date Sep 14, 2023
- Microsoft Technology Licensing, LLC
- Yashesh GAUR
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Application
TRAINING AND USING A TRANSCRIPT GENERATION MODEL ON A MULTI-SPEAKER...
- Publication number 20230215439
- Publication date Jul 6, 2023
- Microsoft Technology Licensing, LLC
- Naoyuki KANDA
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
SEQUENCE-TO-SEQUENCE SPEECH RECOGNITION WITH LATENCY THRESHOLD
- Publication number 20230154467
- Publication date May 18, 2023
- Microsoft Technology Licensing, LLC
- Yashesh GAUR
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
HYPOTHESIS STITCHER FOR SPEECH RECOGNITION OF LONG-FORM AUDIO
- Publication number 20230154468
- Publication date May 18, 2023
- Microsoft Technology Licensing, LLC
- Naoyuki KANDA
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
HYPOTHESIS STITCHER FOR SPEECH RECOGNITION OF LONG-FORM AUDIO
- Publication number 20220199091
- Publication date Jun 23, 2022
- Microsoft Technology Licensing, LLC
- Naoyuki KANDA
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
INTERNAL LANGUAGE MODEL FOR E2E MODELS
- Publication number 20220139380
- Publication date May 5, 2022
- Microsoft Technology Licensing, LLC
- Zhong MENG
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Application
SPEAKER ADAPTATION FOR ATTENTION-BASED ENCODER-DECODER
- Publication number 20220130376
- Publication date Apr 28, 2022
- Microsoft Technology Licensing, LLC
- Zhong MENG
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
DYNAMIC GRADIENT AGGREGATION FOR TRAINING NEURAL NETWORKS
- Publication number 20220036178
- Publication date Feb 3, 2022
- Microsoft Technology Licensing, LLC
- Dimitrios B. DIMITRIADIS
- G06 - COMPUTING CALCULATING COUNTING
Information Patent Application
SEQUENCE-TO-SEQUENCE SPEECH RECOGNITION WITH LATENCY THRESHOLD
- Publication number 20210312923
- Publication date Oct 7, 2021
- Microsoft Technology Licensing, LLC
- Yashesh GAUR
- G10 - MUSICAL INSTRUMENTS ACOUSTICS
Information Patent Application
SPEAKER ADAPTATION FOR ATTENTION-BASED ENCODER-DECODER
- Publication number 20210065683
- Publication date Mar 4, 2021
- Microsoft Technology Licensing, LLC
- Zhong MENG
- G10 - MUSICAL INSTRUMENTS ACOUSTICS

Yashesh GAUR

Person

Patents Grantslast 30 patents

Dynamic gradient aggregation for training neural networks

Training and using a transcript generation model on a multi-speaker...

Hypothesis stitcher for speech recognition of long-form audio

Speaker adaptation for attention-based encoder-decoder

Hypothesis stitcher for speech recognition of long-form audio

Sequence-to-sequence speech recognition with latency threshold

Internal language model for E2E models

Speaker adaptation for attention-based encoder-decoder

Patents Applicationslast 30 patents

TRAINING AND USING A TRANSCRIPT GENERATION MODEL ON A MULTI-SPEAKER...

HYPOTHESIS STITCHER FOR SPEECH RECOGNITION OF LONG-FORM AUDIO

ON-DEVICE STREAMING INVERSE TEXT NORMALIZATION (ITN)

TRAINING AND USING A TRANSCRIPT GENERATION MODEL ON A MULTI-SPEAKER...

SEQUENCE-TO-SEQUENCE SPEECH RECOGNITION WITH LATENCY THRESHOLD

HYPOTHESIS STITCHER FOR SPEECH RECOGNITION OF LONG-FORM AUDIO

HYPOTHESIS STITCHER FOR SPEECH RECOGNITION OF LONG-FORM AUDIO

INTERNAL LANGUAGE MODEL FOR E2E MODELS

SPEAKER ADAPTATION FOR ATTENTION-BASED ENCODER-DECODER

DYNAMIC GRADIENT AGGREGATION FOR TRAINING NEURAL NETWORKS

SEQUENCE-TO-SEQUENCE SPEECH RECOGNITION WITH LATENCY THRESHOLD

SPEAKER ADAPTATION FOR ATTENTION-BASED ENCODER-DECODER