Inventor · White Plains, NY, US

Samuel Thomas

42Patents
8h-index
37Co-inventors
75Inventor score

Filing activity: Nov 6, 1978 → Mar 31, 2023

Most-cited inventions

PatentTitleAreaCited byStatus
US4213486A Coated pipe and process for making same Emerging Cross-Sectional Technologies 45 Expired
US4523141A Pipe coating Physics 19 Expired
US4268334A Pipe wrapping adhesive tape method Emerging Cross-Sectional Technologies 14 Expired
US4589275A Pipe coating Physics 13 Expired
US9697823B1 Acoustic model training Physics 11 Active
US9584758B1 Combining installed audio-visual sensors with ad-hoc mobile audio-visual sensors for smart meeting rooms Electricity 10 Active
US5352292A Apparatus for coating exterior weld joints of a pipe Emerging Cross-Sectional Technologies 10 Expired
US4780158A Novel method for heating cylindrical articles and apparatus useful therefor Performing Operations; Transporting 9 Expired
US9697835B1 Acoustic model training Physics 6 Active
US10380484B2 Annealed dropout training of neural networks Physics 5 Active
US10546575B2 Using recurrent neural network for partitioning of audio data into segments that each correspond to a speech feature cluster identifier Physics 4 Active
US10347270B2 Denoising a signal Physics 3 Active
US10249292B2 Using long short-term memory recurrent neural network for speaker diarization segmentation Physics 3 Active
US10373054B2 Annealed dropout training of neural networks Physics 2 Active
US10629221B2 Denoising a signal Physics 1 Active
US7426631B2 Methods and systems for storing branch information in an address table of a processor Physics 1 Expired
US11183194B2 Detecting and recovering out-of-vocabulary words in voice-to-text transcription systems Physics 1 Active
US10096315B2 Acoustic model training Physics 1 Active
US11908454B2 Integrating text inputs for training and adapting neural network transducer ASR models Physics 1 Active
US9959887B2 Multi-pass speech activity detection strategy to improve automatic speech recognition Physics 1 Active
US11250872B2 Using closed captions as parallel training data for customization of closed captioning systems Electricity 1 Active
US10657980B2 Denoising a signal Physics 1 Active
US11929062B2 End-to-end spoken language understanding without full transcripts Physics 1 Active
US9912909B2 Combining installed audio-visual sensors with ad-hoc mobile audio-visual sensors for smart meeting rooms Electricity 1 Active
US11907845B2 Training teacher machine learning models using lossless and lossy branches Physics 1 Active

Source: USPTO / EPO open patent data. Inventor disambiguation is heuristic; counts are objective bibliographic measures.