Deeksha Dixit

Title: Deeksha Dixit and Occluded Object Detection for Vehicles

Introduction

Deeksha Dixit is an inventor based in San Diego, California. She is associated with Qualcomm Incorporated and is named on 1 patent concerning image processing for vehicle driving assistance systems.

Latest Patents

Dixit’s patent, “Occluded Object Detection and Correction for Vehicle Applications,” describes a method for improving how vehicle systems perceive objects that are partly or fully obscured. The approach uses images captured from different viewpoints to identify portions depicting an object and generate additional views, including views not present in the original images. A machine-learning model then uses corresponding unobstructed views to correct views in which the object is occluded, supporting the creation of a top-view image.

This technology is intended to improve visibility for autonomous driving systems when objects are partially obstructed, helping vehicle systems interpret their surroundings more effectively.

Career Highlights

Dixit’s listed work connects her to Qualcomm Incorporated and to the development of image-processing technology for vehicle applications. Her patent focuses on using multiple viewpoints and machine learning to address occlusion in driving-assistance systems.

Collaborations

The provided information identifies Varun Ravi Kumar and Senthil Kumar Yogamani as Dixit’s coworkers. Their names are associated with her professional context, though specific roles in the patent are not provided.

Conclusion

Deeksha Dixit’s patent addresses a practical challenge in vehicle perception by using machine learning to reconstruct clearer views of occluded objects, contributing to technologies designed to support autonomous driving.