Yan Liu

Title: Yan Liu: Innovator in Image Modeling and Classification

Introduction

Yan Liu is a prominent inventor based in Grenoble, France. She has made significant contributions to the fields of image modeling and classification, holding a total of 4 patents. Her work focuses on developing systems and methods that enhance the representation and classification of images.

Latest Patents

One of her latest patents is titled "Modeling images as mixtures of image models." This patent describes a system and method for generating an image representation by modeling the image as a set of mixture weights. Each weight corresponds to a reference image model, such as Gaussian mixture models (GMMs). The weights are optimized through an objective function that associates each reference image model with its respective weight.

Another notable patent is "Training a classifier by dimension-wise embedding of training data." This invention discloses a method and apparatus for training a linear classifier. In this process, signatures for a set of training samples, such as images, are converted from a first multi-dimensional space to a second multi-dimensional space using embedding functions. The trained linear classifier can approximate the accuracy of a non-linear classifier while reducing computational costs during the learning phase.

Career Highlights

Yan Liu is currently employed at Xerox Corporation, where she continues to innovate in her field. Her work has garnered attention for its practical applications in image processing and machine learning.

Collaborations

Throughout her career, Yan has collaborated with notable colleagues, including Florent C Perronnin and Jorge Sanchez. These collaborations have contributed to her success and the advancement of her research.

Conclusion

Yan Liu is a remarkable inventor whose work in image modeling and classification has led to significant advancements in technology. Her contributions continue to influence the field and inspire future innovations.