Aseem Rastogi

Title: Aseem Rastogi: Innovator in Privacy-Preserving Machine Learning

Introduction

Aseem Rastogi is a notable inventor based in Bengaluru, India. He has made significant contributions to the field of machine learning, particularly in the area of privacy-preserving technologies. His work focuses on enhancing the security of deep neural networks, which are widely used in various applications today.

Latest Patents

Aseem Rastogi holds a patent for "Private inference in deep neural network." This innovation provides a secure inference method over Deep Neural Networks (DNNs) using secure two-party computation. The technology enables privacy-preserving machine learning by employing a specific type of comparison that can be utilized as a building block for various layers in the DNN, including ReLU activations and divisions. The secure inference computes a Boolean share of a bit that indicates whether input value x is less than input value y, where x is held by a user and y is held by a provider of the DNN. This method is more efficient than processing individual bits, making it better suited for complex DNNs.

Career Highlights

Aseem Rastogi is currently associated with Microsoft Technology Licensing, LLC, where he continues to work on innovative solutions in the field of machine learning. His expertise in secure computation and deep learning has positioned him as a valuable asset in the tech industry.

Collaborations

Aseem collaborates with talented individuals such as Nishanth Chandran and Divya Gupta, who contribute to his projects and research endeavors.

Conclusion

Aseem Rastogi's work in privacy-preserving machine learning represents a significant advancement in the field of artificial intelligence. His innovative patent demonstrates the potential for secure applications of deep neural networks, paving the way for future developments in this critical area.