Jiahui Yu
15Patents
4h-index
30Co-inventors
52Inventor score
Filing activity: Jul 26, 2017 → Sep 27, 2024
Most-cited inventions
| Patent | Title | Area | Cited by | Status |
|---|---|---|---|---|
| US10769493B2 | Method and apparatus for neural network training and construction and method and apparatus for object detection | Physics | 36 | Active |
| US10614557B2 | Digital image completion using deep learning | Physics | 11 | Active |
| US10755391B2 | Digital image completion by learning generation and patch matching jointly | Physics | 5 | Active |
| US10672164B2 | Predicting patch displacement maps using a neural network | Physics | 4 | Active |
| US11250548B2 | Digital image completion using deep learning | Physics | 3 | Active |
| US11334971B2 | Digital image completion by learning generation and patch matching jointly | Physics | 2 | Active |
| US10839575B2 | User-guided image completion with image completion neural networks | Physics | 1 | Active |
| US12154581B2 | Cascaded encoders for simplified streaming and non-streaming ASR | Physics | 0 | Active |
| US12190869B2 | Optimizing inference performance for conformer | Physics | 0 | Active |
| US12079703B2 | Convolution-augmented transformer models | Physics | 0 | Active |
| US12282857B1 | Relative margin for contrastive learning | Physics | 0 | Active |
| US12094453B2 | Fast emit low-latency streaming ASR with sequence-level emission regularization utilizing forward and backward probabilities between nodes of an alignment lattice | Physics | 0 | Active |
| US11436775B2 | Predicting patch displacement maps using a neural network | Physics | 0 | Active |
| US12322383B2 | Predicting word boundaries for on-device batching of end-to-end speech recognition models | Physics | 0 | Active |
| US12373666B2 | Convolution-augmented transformer models | Physics | 0 | Active |
Source: USPTO / EPO open patent data. Inventor disambiguation is heuristic; counts are objective bibliographic measures.