Otkrist Gupta

Title: Otkrist Gupta: Innovator in Deep Learning and Data Privacy

Introduction

Otkrist Gupta is a prominent inventor based in Cambridge, MA, known for his significant contributions to the field of deep learning and data privacy. With a total of three patents to his name, Gupta is at the forefront of developing innovative solutions that enhance the security and efficiency of neural networks.

Latest Patents

Gupta's latest patents include groundbreaking work on the secure training of multi-party deep neural networks. This invention allows a deep neural network to be trained using data from multiple entities, referred to as Alices, while ensuring that an external computing entity, known as Bob, can assist in computations without accessing Alices' data. The method employs a 'split' neural network architecture, which consists of an Alice part with at least three neural layers and a Bob part with at least two layers. This design preserves data privacy during the training process.

Another notable patent focuses on methods and apparatus for reducing leakage in distributed deep learning. This invention aims to prevent attackers from reconstructing raw data from activation outputs of intermediate layers in the network. By optimizing the loss function to minimize distance correlation between raw data and activation outputs, Gupta's approach ensures attribute-level privacy while maintaining the integrity of the learning process.

Career Highlights

Otkrist Gupta is affiliated with the Massachusetts Institute of Technology, where he continues to push the boundaries of research in artificial intelligence and machine learning. His work has garnered attention for its practical applications in enhancing data security and privacy in various technological domains.

Collaborations

Gupta collaborates with esteemed colleagues such as Ramesh Raskar and Praneeth Vepakomma, contributing to a dynamic research environment that fosters innovation and creativity in the field of deep learning.

Conclusion

Otkrist Gupta's contributions to deep learning and data privacy exemplify the impact of innovative thinking in technology. His patents reflect a commitment to advancing the field while addressing critical issues related to data security.