Inventor · disambiguated record
Jiahui Yu
Also filed as: YU JIAHUI
16 granted patents·11 pending applications·82 citations·filing 2017–2025
90Inventor score
Top patents by PatentIndex Score
27 records- 0195US10769493B2Method and apparatus for neural network training and construction and method and apparatus for object detectionBEIJING KUANGSHI TECH CO LTD·Filed 2017·Granted Sep 8, 2020·44 cites·7 claims
- 0294US10614557B2Digital image completion using deep learningADOBE INC·Filed 2017·Granted Apr 7, 2020·15 cites·20 claims
- 0393US11334971B2Digital image completion by learning generation and patch matching jointlyADOBE INC·Filed 2020·Granted May 17, 2022·3 cites·20 claims
- 0493US11250548B2Digital image completion using deep learningADOBE INC·Filed 2020·Granted Feb 15, 2022·4 cites·20 claims
- 0591US10755391B2Digital image completion by learning generation and patch matching jointlyADOBE INC·Filed 2018·Granted Aug 25, 2020·10 cites·20 claims
- 0686US10672164B2Predicting patch displacement maps using a neural networkADOBE INC·Filed 2017·Granted Jun 2, 2020·4 cites·20 claims
- 0778US12373666B2Convolution-augmented transformer modelsGOOGLE LLC·Filed 2024·Granted Jul 29, 2025·0 cites·20 claims
- 0877US2025335739A1Convolution-Augmented Transformer ModelsGOOGLE LLC·Filed 2025·Application pending·0 cites
- 0974US12282857B1Relative margin for contrastive learningGOOGLE LLC·Filed 2024·Granted Apr 22, 2025·0 cites·18 claims
- 1072US10839575B2User-guided image completion with image completion neural networksADOBE INC·Filed 2018·Granted Nov 17, 2020·2 cites·20 claims
- 1170US2024112088A1Vector-Quantized Image ModelingGOOGLE LLC·Filed 2023·Application pending·0 cites
- 1269US2025217656A1Relative margin for contrastive learningGOOGLE LLC·Filed 2025·Application pending·0 cites
- 1365US12079703B2Convolution-augmented transformer modelsGOOGLE LLC·Filed 2020·Granted Sep 3, 2024·0 cites·18 claims
- 1465US2024404238A1Vector-Quantized Image ModelingGOOGLE LLC·Filed 2022·Application pending·0 cites
- 1562US11436775B2Predicting patch displacement maps using a neural networkADOBE INC·Filed 2020·Granted Sep 6, 2022·0 cites·20 claims
- 1661US12482455B2Systems and methods for training dual-mode machine-learned speech recognition modelsGOOGLE LLC·Filed 2021·Granted Nov 25, 2025·0 cites·21 claims
- 1755US2023351149A1Contrastive captioning neural networksGOOGLE LLC·Filed 2023·Application pending·0 cites
- 1855US2025054322A1Attribute Recognition with Image-Conditioned Prefix Language ModelingGOOGLE LLC·Filed 2024·Application pending·0 cites
- 1954US12154581B2Cascaded encoders for simplified streaming and non-streaming ASRGOOGLE LLC·Filed 2021·Granted Nov 26, 2024·0 cites·17 claims
- 2054US12094453B2Fast emit low-latency streaming ASR with sequence-level emission regularization utilizing forward and backward probabilities between nodes of an alignment latticeGOOGLE LLC·Filed 2021·Granted Sep 17, 2024·0 cites·24 claims
- 2154US2025111671A1Media item characterization based on multimodal embeddingsGOOGLE LLC·Filed 2024·Application pending·0 cites
- 2253US12190869B2Optimizing inference performance for conformerGOOGLE LLC·Filed 2022·Granted Jan 7, 2025·0 cites·18 claims
- 2353US2025037426A1Co-Training of Action Recognition Machine Learning ModelsGOOGLE LLC·Filed 2022·Application pending·0 cites
- 2449US12322383B2Predicting word boundaries for on-device batching of end-to-end speech recognition modelsGOOGLE LLC·Filed 2022·Granted Jun 3, 2025·0 cites·26 claims
- 2549US2025124708A1Video-text modeling with zero-shot transfer from contrastive captionersGOOGLE LLC·Filed 2023·Application pending·0 cites
- 2648US2025118291A1Self-supervised learning for audio processingGOOGLE LLC·Filed 2023·Application pending·0 cites
- 2746US2023281400A1Systems and Methods for Pretraining Image Processing ModelsGOOGLE LLC·Filed 2022·Application pending·0 cites
Join the waitlist — get patent alerts
Get an alert when Jiahui Yu 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 →