Eric A. Sather
39Patents
6h-index
14Co-inventors
66Inventor score
Filing activity: Dec 27, 2002 → Aug 27, 2023
Most-cited inventions
| Patent | Title | Area | Cited by | Status |
|---|---|---|---|---|
| US8352296B2 | Managing real time meeting room status | Physics | 62 | Active |
| US10592732B1 | Probabilistic loss function for training network with triplets | Physics | 21 | Active |
| US11610154B1 | Preventing overfitting of hyperparameters during training of network | Physics | 14 | Active |
| US9300912B2 | Software based whiteboard capture solution for conference room meetings | Electricity | 11 | Active |
| US11531879B1 | Iterative transfer of machine-trained network inputs from validation set to training set | Physics | 6 | Active |
| US10671888B1 | Using batches of training items for training a network | Physics | 6 | Active |
| US8789001B1 | System and method for using fabric-graph flow to determine resource costs | Physics | 6 | Active |
| US7385420B1 | Repeatable block producing a non-uniform routing architecture in a field programmable gate array having segmented tracks | Electricity | 4 | Active |
| US6891396B1 | Repeatable block producing a non-uniform routing architecture in a field programmable gate array having segmented tracks | Electricity | 4 | Expired |
| US11113603B1 | Training network with discrete weight values | Physics | 4 | Active |
| US11017295B1 | Device storing ternary weight parameters for machine-trained network | Physics | 3 | Active |
| US11151695B1 | Video denoising using neural networks with spatial and temporal features | Physics | 3 | Active |
| US11604973B1 | Replication of neural network layers | Physics | 3 | Active |
| US11163986B2 | Using batches of training items for training a network | Physics | 3 | Active |
| US11586902B1 | Training network to minimize worst case surprise | Physics | 3 | Active |
| US11847567B1 | Loss-aware replication of neural network layers | Physics | 3 | Active |
| US11900238B1 | Removing nodes from machine-trained network based on introduction of probabilistic noise during training | Physics | 2 | Active |
| US7075334B1 | Repeatable block producing a non-uniform routing architecture in a field programmable gate array having segmented tracks | Electricity | 2 | Expired |
| US11537870B1 | Training sparse networks with discrete weight values | Electricity | 2 | Active |
| US11429861B1 | Device storing multiple sets of parameters for machine-trained network | Physics | 2 | Active |
| US11741369B2 | Using batches of training items for training a network | Physics | 2 | Active |
| US12061988B1 | Decomposition of ternary weight tensors | Physics | 1 | Active |
| US12136039B1 | Optimizing global sparsity for neural network | Physics | 1 | Active |
| US9047431B2 | System and method for using fabric-graph flow to determine resource costs | Physics | 1 | Active |
| US12061981B1 | Decomposition of weight tensors in network with value quantization | Physics | 1 | Active |
Source: USPTO / EPO open patent data. Inventor disambiguation is heuristic; counts are objective bibliographic measures.