IDiyas

For the Inventor, By the Inventor

White Plains, NY, United States of America

Jialong Zhang

LinkedIn
USPTO Granted Patents = 20 

Average Co-Inventor Count = 4.9

ph-index = 5

Forward Citations = 184(Granted Patents)


Location History:

  • College Station, TX (US) (2015 - 2017)
  • San Jose, CA (US) (2022 - 2023)
  • White Plains, NY (US) (2020 - 2024)

Company Filing History:

goldMedal19 out of 164,108 
 
International Business Machines Corporation
 patents
silverMedal1 out of 81 
 
Narus, Inc.
 patents

Years Active: 2015-2024

where 'Filed Patents' based on already Granted Patents

20 patents (USPTO):
1. 12045713 - Detecting adversary attacks on a deep neural network (DNN) (2024-07-23),
2. 11886989 - System for measuring information leakage of deep learning models (2024-01-30),
3. 11829879 - Detecting adversarial attacks through decoy training (2023-11-28),
4. 11816575 - Verifiable deep learning training service (2023-11-14),
5. 11775637 - Recurrent neural network based anomaly detection (2023-10-03),
6. 11632393 - Detecting and mitigating malware by evaluating HTTP errors (2023-04-18),
7. 11501156 - Detecting adversarial attacks through decoy training (2022-11-15),
8. 11443182 - Privacy enhancing deep learning cloud service using a trusted execution environment (2022-09-13),
9. 11301563 - Recurrent neural network based anomaly detection (2022-04-12),
10. 11188789 - Detecting poisoning attacks on neural networks by activation clustering (2021-11-30),
11. 11184374 - Endpoint inter-process activity extraction and pattern matching (2021-11-23),
12. 11163860 - Protecting deep learning models using watermarking (2021-11-02),
13. 11144642 - Deactivating evasive malware (2021-10-12),
14. 11132444 - Using gradients to detect backdoors in neural networks (2021-09-28),
15. 10733292 - Defending against model inversion attacks on neural networks (2020-08-04),

Title: The Innovative Contributions of Jialong Zhang in Deep Learning and Neural Networks

Introduction: Jialong Zhang, based in White Plains, NY, has made significant strides in the field of artificial intelligence, particularly in deep learning and neural network security. With a remarkable portfolio of 20 patents, Zhang's work focuses on safeguarding deep learning models from adversarial attacks and information leakage.

Latest Patents: Among Zhang's latest patents is a method for detecting adversary attacks on deep neural networks (DNN). This innovation involves receiving a training data set and during the training of the DNN, recording a representation of activations associated with its intermediate layers. Subsequently, separate classifiers are trained to develop an outlier detection model that identifies adversarial inputs. An action is initiated to protect the DNN deployed system if such an input is detected. Another notable patent highlights a system for measuring information leakage within deep learning models. Here, similarities among intermediate representations are analyzed to determine a partitioning point for the DNN layers, ensuring that one partition of the system does not leak sensitive information.

Career Highlights: Zhang has worked with leading technology companies, including the International Business Machines Corporation (IBM) and Narus, Inc. His experience in these organizations has contributed to his deep understanding of the complexities involved in developing secure AI systems.

Collaborations: Throughout his career, Zhang has collaborated with notable professionals in the field, such as Zhongshu Gu and Heqing Huang. These partnerships have enabled a synergistic approach to innovation, fostering the development of cutting-edge technologies.

Conclusion: Jialong Zhang's dedication to advancing deep learning technologies and enhancing their security reflects his innovative spirit and expertise. With his continuous contributions shaped by both individual ingenuity and collaborative efforts, he stands out as a significant inventor in the modern technological landscape.

This text is generated by artificial intelligence and may not be accurate.
Please report any incorrect information to support@idiyas.com
Loading…