Vahdat Abdelzad

Title: Vahdat Abdelzad and Advances in 3D Perception

Introduction

Vahdat Abdelzad is an inventor based in Waterloo, Canada, whose work includes methods for generating training data and three-dimensional representations from point-cloud data. He is associated with Huawei Technologies Canada Co., Ltd. His listed portfolio includes 2 patents.

Latest Patents

“Method and a Server for Generating Training Data for Training an Object Detector” describes a system for creating training data to fine-tune an object detector. It selects a quasi-stationary bounding box from a sequence of labelled 3D point clouds, relocates an object’s bounding box to a corresponding position in another point cloud, and stores the relabelled data. This approach supports the preparation of training examples from source-domain data.

“Methods and Systems for Generating a Three-Dimensional Representation of a Scene” concerns building 3D scene representations from successive point clouds. The method detects an object, estimates its position in a later point cloud, and uses the object’s dynamic characteristics to determine how many preceding point clouds to aggregate. The resulting augmented point cloud may support more accurate analysis of driving scenarios.

Career Highlights

Abdelzad’s listed inventions focus on 3D point-cloud processing, object detection, and scene representation. Together, they address data preparation for machine-learning systems and the aggregation of observations across time.

Collaborations

Abdelzad is associated with Huawei Technologies Canada Co., Ltd. His listed coworkers include Chengjie Huang and Sean Albert Sedwards.

Conclusion

Vahdat Abdelzad’s patent work centers on techniques for improving object-detector training data and generating 3D representations from point clouds, with potential applications in interpreting driving scenarios.