Kawtar Zaher

Title: Kawtar Zaher: Innovator in 3D Object Generation

Introduction

Kawtar Zaher is a notable inventor based in Vélizy-Villacoublay, France. He has made significant contributions to the field of machine learning, particularly in the generation of 3D objects. His innovative approach has led to the development of a unique patent that enhances the capabilities of 3D modeling.

Latest Patents

Kawtar Zaher holds a patent for a "Canonicalized codebook for 3D object generation." This computer-implemented method of machine learning involves obtaining a training dataset of 3D models of real-world objects. The method includes learning from the training dataset and a patch-decomposition of the 3D models to create a finite codebook of quantized vectors and a neural network. The neural network features a rotation-invariant encoder designed for encoding patches of 3D models into quantized latent vectors. Additionally, it includes a decoder that translates sequences of quantized latent vectors back into 3D models. This innovative solution significantly improves the process of 3D model generation.

Career Highlights

Throughout his career, Kawtar Zaher has worked with prominent organizations such as Dassault Systèmes and École Polytechnique. His experience in these institutions has allowed him to refine his skills and contribute to cutting-edge research in 3D modeling and machine learning.

Collaborations

Kawtar has collaborated with talented individuals, including Mariem Mezghanni and Malika Boulkenafed. These partnerships have fostered a creative environment that encourages innovation and the sharing of ideas.

Conclusion

Kawtar Zaher is a pioneering inventor whose work in 3D object generation through machine learning is making a significant impact in the field. His patent and career achievements reflect his dedication to advancing technology and innovation.