Isht Dwivedi

Title: Isht Dwivedi: Innovator in Action Segmentation Technologies

Introduction

Isht Dwivedi is a prominent inventor based in Santa Clara, CA, known for his contributions to the field of action segmentation technologies. With a total of 5 patents to his name, Dwivedi has made significant strides in developing systems that enhance the understanding of actions in video data.

Latest Patents

One of his latest patents is titled "System and method for providing weakly-supervised online action segmentation." This innovative system involves receiving image data from multi-view videos of a procedure that includes various atomic actions. The method analyzes this image data using weakly-supervised action segmentation to identify each atomic action through an ordered sequence of action labels. Additionally, it trains a neural network with data related to these actions, enabling real-time online action segmentation.

Another notable patent is "Systems and methods for bird's eye view segmentation." This method begins with receiving an input image from an image sensor on an agent. It extracts features from the input image and estimates a depth map for the pixels. The process generates a 3D point map and voxel grid, ultimately leading to the creation of a feature map that facilitates BEV segmentation.

Career Highlights

Isht Dwivedi is currently employed at Honda Motor Co., Ltd., where he applies his expertise in action segmentation technologies. His work focuses on enhancing the capabilities of systems that interpret complex video data, contributing to advancements in various applications.

Collaborations

Throughout his career, Dwivedi has collaborated with notable colleagues, including Behzad Dariush and Athmanarayanan Lakshmi Narayanan. These collaborations have fostered innovation and the development of cutting-edge technologies in the field.

Conclusion

Isht Dwivedi's work in action segmentation technologies exemplifies the impact of innovative thinking in the realm of video data analysis. His patents reflect a commitment to advancing technology and improving real-time action recognition systems.