Ryan Paul Thomas
19Patents
10h-index
29Co-inventors
64Inventor score
Filing activity: Mar 19, 2012 → Sep 23, 2019
Most-cited inventions
| Patent | Title | Area | Cited by | Status |
|---|---|---|---|---|
| US9070366B1 | Architecture for multi-domain utterance processing | Physics | 410 | Active |
| US9275637B1 | Wake word evaluation | Physics | 240 | Active |
| US9697827B1 | Error reduction in speech processing | Physics | 228 | Active |
| US8983844B1 | Transmission of noise parameters for improving automatic speech recognition | Physics | 202 | Active |
| US9818407B1 | Distributed endpointing for speech recognition | Physics | 162 | Active |
| US8898064B1 | Identifying candidate passwords from captured audio | Physics | 137 | Active |
| US8996372B1 | Using adaptation data with cloud-based speech recognition | Physics | 77 | Active |
| US9706406B1 | Security measures for an electronic device | Electricity | 51 | Active |
| US9047868B1 | Language model data collection | Physics | 29 | Active |
| US9286897B2 | Speech recognizer with multi-directional decoding | Electricity | 18 | Active |
| US9436678B2 | Architecture for multi-domain natural language processing | Physics | 9 | Active |
| US10956132B1 | Unified code and data management for model development | Physics | 8 | Active |
| US10424292B1 | System for recognizing and responding to environmental noises | Physics | 7 | Active |
| US9754589B2 | Architecture for multi-domain natural language processing | Physics | 7 | Active |
| US10964315B1 | Monophone-based background modeling for wakeword detection | Physics | 4 | Active |
| US11176936B2 | Architecture for multi-domain natural language processing | Physics | 2 | Active |
| US9959869B2 | Architecture for multi-domain natural language processing | Physics | 1 | Active |
| US10283119B2 | Architecture for multi-domain natural language processing | Physics | 0 | Active |
| US11862153B1 | System for recognizing and responding to environmental noises | Physics | 0 | Active |
Source: USPTO / EPO open patent data. Inventor disambiguation is heuristic; counts are objective bibliographic measures.