Hideyuki Winston Inada

Title: Hideyuki Winston Inada: Inventor of Scalable Wavelet Transformer Technology

Introduction

Hideyuki Winston Inada is an inventor based in San Jose, California. His work includes a patent for a method of combining transformer-based feature extraction with wavelet filtering. He is associated with Microsoft Technology Licensing, LLC.

Latest Patents

Inada is named on 1 patent, titled “Scalable Wavelet Transformer-Based Feature Extraction.” The invention describes receiving digital data and generating embedded patches from it. A transformer block processes an embedded patch, after which wavelet filters create filtered patches. These are combined to produce training data, which is then used to generate a trained prediction model.

Career Highlights

The patent reflects work on scalable approaches to extracting features from digital data. By combining transformer processing with wavelet filters, the disclosed method provides a pipeline for preparing training data and producing a prediction model.

Collaborations

Inada’s listed coworkers include Haichao Wei and Chengming Jiang. Their names are associated with the collaborative environment surrounding his patent work.

Conclusion

Hideyuki Winston Inada’s patent documents an approach that integrates transformer blocks and wavelet filtering to support feature extraction and prediction-model training.