Biography
Pu Zhao is a Research Assistant Professor in Electrical and Computer Engineering, Northeastern University. His research interests include
- Efficient deep learning with weight pruning and neural architecture search to obtain efficient DNNs with less parameters & computations, and achieve real-time DNN inference on resource-limited mobile phones
- Evaluation of the robustness of DNNs from an optimization perspective (such as exploration of adversarial attack methods to craft adversarial examples and investigation of defending methods against adversarial attacks)
Download my resumé.
Interests
- Artificial Intelligence
- Adversarial Robustness
- Model pruning
- Real-Time Deployment
Education
PhD in Electrical and Computer Engineering, 2021
Northeastern University
MEng in Electrical Engineering, 2016
Shanghai Jiao Tong University
BSc in Electrical Engineering, 2013
Shanghai Jiao Tong University
Experience
Research Assistant Professor
Northeastern University
Research Internship
Explore the image search with text assistance.
Research Fall Internship-Graduate
IBM Thomas J. Watson Research Center
Investigate the defense performance of model connection against three kinds of attacks including backdoor attack, fault injection attack and adversarial attack.
Teaching Assistance
Computer Architecture
Help with the teaching of computer architecture such as designing the course project and answering questions during office hours.
Accomplishments
First Place in ISLPED 2020 Design Contest
CoCoPIE: A Framework of Compression-Compilation Co-design Towards Ultra-High Energy Efficiency and Real Time DNN Inference on Mobile Devices
See certificateBest Paper Nomination in ICCAD 2018
Defensive Dropout for Hardening Deep Neural Networks under Adversarial Attacks
National Scholarship for Postgraduate Students
National Scholarship for Postgraduate Students
Academic Excellence Scholarship for Postgraduate Students
Academic Excellence Scholarship for Postgraduate Students
Academic Excellence Scholarship
Academic Excellence Scholarship
First Prize at 26th National Physics Contest for College Students
Featured Publications
Towards Real-Time DNN Inference on Mobile Platforms with Model Pruning and Compiler Optimization
High-end mobile platforms rapidly serve as primary computing devices for a wide range of Deep Neural Network (DNN) applications. …
CoCoPIE: Enabling Real-Time AI on Off-the-Shelf Mobile Devices via Compression-Compilation Co-Design
Assuming hardware is the major constraint for enabling real-time mobile intelligence, the industry has mainly dedicated their efforts …
Recent Publications
Quickly discover relevant content by filtering publications.
(2023).
Towards Real-Time Segmentation on the Edge.
In AAAI.
(2022).
Advancing Model Pruning via Bi-level Optimization.
In NeurIPS.
(2022).
All-in-One: A Highly Representative DNN Pruning Framework for Edge Devices with Dynamic Power Management.
In ICCAD.
(2022).
Compiler-Aware Neural Architecture Search for On-Mobile Real-time Super-Resolution.
In ECCV.
Contact
- zhao.pu@northeastern.edu
- 140 the Fenway, Boston, MA 02120
- Enter Building and take the stairs to Office R306 on Floor 3
- Book an appointment
- DM Me
- Skype Me
- Chat on Keybase
- Discuss on Forum