Inventor · disambiguated record
Hasim Sak
Also filed as: SAK HASIM
40 granted patents·6 pending applications·612 citations·filing 2013–2024
97Inventor score
Top patents by PatentIndex Score
46 records- 0198US10229672B1Training acoustic models using connectionist temporal classificationGOOGLE LLC·Filed 2017·Granted Mar 12, 2019·59 cites·20 claims
- 0297US9786270B2Generating acoustic modelsGOOGLE INC·Filed 2016·Granted Oct 10, 2017·76 cites·20 claims
- 0397US9292489B1Sub-lexical language models with word level pronunciation lexiconsGOOGLE INC·Filed 2013·Granted Mar 22, 2016·252 cites·23 claims
- 0496US9460088B1Written-domain language modeling with decompositionGOOGLE INC·Filed 2013·Granted Oct 4, 2016·48 cites·22 claims
- 0594US11769493B2Training acoustic models using connectionist temporal classificationGOOGLE LLC·Filed 2022·Granted Sep 26, 2023·2 cites·20 claims
- 0694US10706840B2Encoder-decoder models for sequence to sequence mappingGOOGLE LLC·Filed 2017·Granted Jul 7, 2020·19 cites·17 claims
- 0794US9728185B2Recognizing speech using neural networksGOOGLE INC·Filed 2015·Granted Aug 8, 2017·23 cites·18 claims
- 0893US11741947B2Transformer transducer: one model unifying streaming and non-streaming speech recognitionGOOGLE LLC·Filed 2021·Granted Aug 29, 2023·3 cites·27 claims
- 0993US9818410B2Speech recognition with acoustic modelsGOOGLE INC·Filed 2015·Granted Nov 14, 2017·12 cites·20 claims
- 1092US9818409B2Context-dependent modeling of phonemesGOOGLE INC·Filed 2015·Granted Nov 14, 2017·13 cites·17 claims
- 1192US9620108B2Processing acoustic sequences using long short-term memory (LSTM) neural networks that include recurrent projection layersGOOGLE INC·Filed 2014·Granted Apr 11, 2017·29 cites·20 claims
- 1290US11961515B2Contrastive Siamese network for semi-supervised speech recognitionGOOGLE LLC·Filed 2021·Granted Apr 16, 2024·2 cites·24 claims
- 1390US10803855B1Training acoustic models using connectionist temporal classificationGOOGLE LLC·Filed 2019·Granted Oct 13, 2020·6 cites·26 claims
- 1489US11341958B2Training acoustic models using connectionist temporal classificationGOOGLE LLC·Filed 2020·Granted May 24, 2022·2 cites·18 claims
- 1588US10325602B2Neural networks for speaker verificationGOOGLE LLC·Filed 2017·Granted Jun 18, 2019·12 cites·20 claims
- 1687US12266347B2End-to-end multi-talker overlapping speech recognitionGOOGLE LLC·Filed 2022·Granted Apr 1, 2025·1 cites·16 claims
- 1787US10192556B2Speech recognition with acoustic modelsGOOGLE LLC·Filed 2017·Granted Jan 29, 2019·5 cites·20 claims
- 1887US9721562B2Generating representations of acoustic sequencesGOOGLE INC·Filed 2014·Granted Aug 1, 2017·9 cites·20 claims
- 1986US10134393B2Generating representations of acoustic sequencesGOOGLE LLC·Filed 2017·Granted Nov 20, 2018·4 cites·20 claims
- 2086US10127904B2Learning pronunciations from acoustic sequencesGOOGLE LLC·Filed 2015·Granted Nov 13, 2018·8 cites·20 claims
- 2181US10923112B2Generating representations of acoustic sequencesGOOGLE LLC·Filed 2019·Granted Feb 16, 2021·2 cites·20 claims
- 2281US10783900B2Convolutional, long short-term memory, fully connected deep neural networksGOOGLE LLC·Filed 2015·Granted Sep 22, 2020·4 cites·17 claims
- 2380US10535338B2Generating representations of acoustic sequencesGOOGLE LLC·Filed 2018·Granted Jan 14, 2020·2 cites·30 claims
- 2478US11521595B2End-to-end multi-talker overlapping speech recognitionGOOGLE LLC·Filed 2020·Granted Dec 6, 2022·1 cites·22 claims
- 2578US9208779B2Mixture of n-gram language modelsGOOGLE INC·Filed 2013·Granted Dec 8, 2015·7 cites·20 claims
- 2677US11715486B2Convolutional, long short-term memory, fully connected deep neural networksGOOGLE LLC·Filed 2019·Granted Aug 1, 2023·2 cites·10 claims
- 2777US10733979B2Latency constraints for acoustic modelingGOOGLE LLC·Filed 2015·Granted Aug 4, 2020·3 cites·17 claims
- 2876US2023410796A1Encoder-decoder models for sequence to sequence mappingGOOGLE LLC·Filed 2023·Application pending·0 cites
- 2974US12254869B2One model unifying streaming and non-streaming speech recognitionGOOGLE LLC·Filed 2023·Granted Mar 18, 2025·0 cites·18 claims
- 3074US10026397B2Processing acoustic sequences using long short-term memory (LSTM) neural networks that include recurrent projection layersGOOGLE LLC·Filed 2017·Granted Jul 17, 2018·2 cites·20 claims
- 3171US12334059B2Contrastive Siamese network for semi-supervised speech recognitionGOOGLE LLC·Filed 2024·Granted Jun 17, 2025·0 cites·20 claims
- 3271US11721327B2Generating representations of acoustic sequencesGOOGLE LLC·Filed 2021·Granted Aug 8, 2023·0 cites·20 claims
- 3371US2024371379A1Reducing Streaming ASR Model Delay With Self AlignmentGOOGLE LLC·Filed 2024·Application pending·0 cites
- 3470US10431206B2Multi-accent speech recognitionGOOGLE LLC·Filed 2016·Granted Oct 1, 2019·2 cites·20 claims
- 3565US11776531B2Encoder-decoder models for sequence to sequence mappingGOOGLE LLC·Filed 2020·Granted Oct 3, 2023·0 cites·20 claims
- 3663US11996088B2Setting latency constraints for acoustic modelsGOOGLE LLC·Filed 2020·Granted May 28, 2024·0 cites·20 claims
- 3762US9594744B2Speech transcription including written textGOOGLE INC·Filed 2013·Granted Mar 14, 2017·2 cites·20 claims
- 3861US12482470B2Speaker-turn-based online speaker diarization with constrained spectral clusteringGOOGLE LLC·Filed 2021·Granted Nov 25, 2025·0 cites·24 claims
- 3961US12057124B2Reducing streaming ASR model delay with self alignmentGOOGLE LLC·Filed 2021·Granted Aug 6, 2024·0 cites·20 claims
- 4057US12315499B2Semi-supervised training scheme for speech recognitionGOOGLE LLC·Filed 2022·Granted May 27, 2025·0 cites·24 claims
- 4154US2024177706A1Monte Carlo Self-Training for Speech RecognitionGOOGLE LLC·Filed 2023·Application pending·0 cites
- 4249US2025378286A1Application Programming Interfaces For On-Device Speech ServicesGOOGLE LLC·Filed 2022·Application pending·0 cites
- 4346US9786269B2Language modeling of complete language sequencesGOOGLE INC·Filed 2013·Granted Oct 10, 2017·0 cites·20 claims
- 4441US10275704B2Generating representations of input sequences using neural networksGOOGLE LLC·Filed 2015·Granted Apr 30, 2019·0 cites·12 claims
- 4537US2016035344A1Identifying the language of a spoken utteranceGOOGLE INC·Filed 2015·Application pending·0 cites
- 4635US2018174576A1Acoustic-to-word neural network speech recognizerGOOGLE LLC·Filed 2017·Application pending·0 cites
Join the waitlist — get patent alerts
Get an alert when Hasim Sak files or is granted a new patent.
We store only your email — no account needed. See our privacy policy.
Identity basis: PatentsView inventor disambiguation (2025Q4-odp release). How scoring works →