Ying Wang

Title: Ying Wang and Small-Mask Image Editing.

Introduction

Ying Wang is an inventor based in Bothell, Washington. His work includes image-editing techniques designed to improve inpainting when the selected mask covers only a small portion of an image. Wang is associated with Amazon Technologies, Inc.

Latest Patents

Wang is credited with 1 patent, titled “Image Editing for Small Masks.” The patent describes a process for handling image edits when the mask is small relative to the source image. The method generates a second source image that includes the masked pixels and some surrounding context, then upsamples both the image and mask data. A diffusion-based inpainting model uses the upsampled data to create an inpainted image in the region corresponding to the mask.

Career Highlights

The patent highlights an approach to a practical challenge in image editing: preserving useful context and detail when an edit targets a small area. By expanding and upsampling the relevant image region and mask, the technique aims to support diffusion-based inpainting for localized edits.

Collaborations

Wang’s listed coworker is Shamit Lal. Both are associated with Amazon Technologies, Inc.

Conclusion

Ying Wang’s patent focuses on improving diffusion-based image inpainting for small masks by preparing and upsampling image and mask data, reflecting his contribution to image-editing innovation.