Adaptive Computing and Embedded Systems

The Adaptive Computing and Embedded Systems (ACES) Lab, lead by Prof. Farinaz Koushanfar, focuses on making intelligent data-intensive embedded computing applications and systems. The added intelligence is to satisfy security and robustness, energy-efficiency, timeliness, IP protection rights, design automation, and many more requirements of emerging technologies.

• 

             Obama2 Placeholder for descriptions
  

• 

             TR35 Placeholder for descriptions
  

• 

             Group UCSD Placeholder for descriptions
  

Our modern society is increasingly dependent on embedded computing devices that process a vast amount of data. Common examples include visual computing on mobile phones, automotive embedded system, wireless sensor networks, implantable medical devices, and mobile virtual reality games. The content, communication, and processing algorithms can be overwhelming on small platforms. What exacerbates the problem are the real-time constraints set by certain applications, which prohibits outsourcing to the cloud due to the incurred delay and uncertainty.

To be viable, there are at least two major sets of technical challenges that need to be addressed for small form-factor platforms that enable present and pending Internet-of-Everything (IoE) systems. One set of hurdles has to do with resource and/or application constraints such as real-time, available energy, or memory. Another set of barriers arises due to security, reliability or safety requirements. Attacks on these systems go far beyond destruction of data, as they have the potential to impact physical assets and people’s lives. In this context, classic solutions for resource-efficiency and/or security are of limited effectiveness.

A distinguishing characteristic of our research is bridging the divide between the theory and implementation. One the one hand, we use theoretical methods from various disciplines such as statistics, optimization, and cryptography to address complex hardware engineering problems. On the other hand, we utilize practical computer engineering and design automation techniques to address sophisticated problems in security, data mining, and emerging technologies. Simultaneous with development of models and theory, we realize and deploy the results in practice whenever possible. In such a way, we obtain a proof-of-concept from our realization, which often proposes novel methodologies and newer research directions. We also have a number of collaborative projects with experts in relevant fields (e.g., security, data mining, computer architecture).

Recent publications

@article {Scalable23, month = {7}, url = {https://dl.acm.org/doi/pdf/10.1145/3607192}, year = {2023}, author = {Xinqiao Zhang and Mohammad Samragh and Siam Hussain and Ke Huang and Farinaz Koushanfar}, publisher = {ACM Transactions on Embedded Computing Systems}, title = {Scalable Binary Neural Network applications in Oblivious Inference}, _pageName = {Scalable23} } @article {Swnet2021, month = {12}, url = {https://ieeexplore.ieee.org/abstract/document/9600865}, year = {2021}, author = {Mojan Javaheripi and Bita Darvish Rouhani and Farinaz Koushanfar}, publisher = {IEEE Journal on Emerging and Selected Topics in Circuits and Systems}, title = {SWANN: Small-World Architecture for Fast Convergence of Neural Networks}, pdf = {SWANN 2021.pdf}, _pageName = {Swnet2021} } @article {Autorank2021, month = {12}, url = {https://ieeexplore.ieee.org/abstract/document/9611209}, year = {2021}, author = {Mojan Javaheripi and Mohammad Samragh and Farinaz Koushanfar}, publisher = {IEEE Journal on Emerging and Selected Topics in Circuits and Systems}, title = {AutoRank: Automated Rank Selection for Effective Neural Network Customization}, pdf = {AutoRank 2021.pdf}, _pageName = {Autorank2021} } @conference {Hussain2021, month = {11}, url = {https://dl.acm.org/doi/10.1145/3460120.3484797}, year = {2021}, author = {Siam Umar Hussain and Mojan Javaheripi and Mohammad Samragh and Farinaz Koushanfar}, publisher = {ACM SIGSAC Conference on Computer and Communications Security (CCS)}, title = {COINN: Crypto/ML Codesign for Oblivious Inference via Neural Networks}, pdf = {COINN cameraReady.pdf}, _pageName = {Hussain2021} } @conference {Hashtag2021, month = {11}, url = {https://ieeexplore.ieee.org/abstract/document/9643556}, year = {2021}, author = {Mojan Javaheripi and Farinaz Koushanfar}, publisher = {IEEE/ACM International Conference On Computer Aided Design (ICCAD)}, title = {HASHTAG: Hash Signatures for Online Detection of Fault-Injection Attacks on Deep Neural Networks}, pdf = {Hashtag 2021.pdf}, _pageName = {Hashtag2021} } @conference {Chen2021DTRAP, month = {10}, year = {2021}, author = {Huili Chen and Cheng Fu and Jishen Zhao and Farinaz Koushanfar}, publisher = {ACM Digital Threats: Research and Practice (DTRAP)}, title = {GALU: A Genetic Algorithm Framework for Logic Unlocking. ACM Digital Threats: Research and Practice (DTRAP)}, pdf = {2021DTRAP GALU CamVer.pdf}, _pageName = {Chen2021DTRAP} } @inproceedings {Hussain2021USENIX, isbn = {978-1-939133-24-3}, booktitle = {30th {USENIX} Security Symposium ({USENIX} Security 21)}, month = {8}, url = {https://www.usenix.org/conference/usenixsecurity21/presentation/hussain}, pages = {2273--2290}, year = {2021}, author = {Shehzeen Hussain and Paarth Neekhara and Shlomo Dubnov and Julian McAuley and Farinaz Koushanfar}, journal = {USENIX Security}, publisher = {USENIX Association}, title = {WaveGuard: Understanding and Mitigating Audio Adversarial Examples}, pdf = {Usenixsec21-hussain.pdf}, _pageName = {Hussain2021USENIX} } @inproceedings {Javaheripi2021XCL, month = {6}, year = {2021}, author = {Greg Fields and Mohammad Samragh and Mojan Javaheripi and Farinaz Koushanfar and Tara Javidi}, publisher = {IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)}, title = {Trojan Signatures in DNN Weights}, pdf = {Fields Trojan Signatures.pdf}, _pageName = {Javaheripi2021XCL} } @inproceedings {WACV2021, month = {1}, url = {https://openaccess.thecvf.com/content/WACV2021/html/Hussain_Adversarial_Deepfakes_Evaluating_Vulnerability_of_Deepfake_Detectors_to_Adversarial_Examples_WACV_2021_paper.html}, year = {2021}, author = {Hussain, Shehzeen and Neekhara, Paarth and Jere, Malhar and Koushanfar, Farinaz and McAuley, Julian}, publisher = {IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)}, title = {Adversarial Deepfakes: Evaluating Vulnerability of Deepfake Detectors to Adversarial Examples}, _pageName = {WACV2021} } @proceedings {Mohammad2021CVPR, url = {https://openaccess.thecvf.com/content/CVPR2021W/BiVision/papers/Samragh_On_the_Application_of_Binary_Neural_Networks_in_Oblivious_Inference_CVPRW_2021_paper.pdf}, year = {2021}, author = {Mohammad Samragh and Siam Hussain and Xinqiao Zhang and Ke Huang and Farinaz Koushanfar}, publisher = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, title = {On the Application of Binary Neural Networks in Oblivious Inference}, pdf = {Samragh On the Application of Binary Neural Networks in Oblivious Inference CVPRW 2021 paper.pdf}, _pageName = {Mohammad2021CVPR} }

People

Supervisor

Farinaz Koushanfarmail_outline

Principal Coordinator

Postdoctoral

Zahra Ghodsi

Postdoctoral Scholar

Position available

XXX Topic

Position available

XXX Topic

PhD

Huili Chenmail_outline

XXX Topic

Mojan Javaheripimail_outline

XXX Topic

Shehzeen Hussainmail_outlinepublic

XXX Topic

Xinqiao Zhangmail_outlinepublic

XXX Topic

Cody Emersonmail_outline

XXX Topic

Nojan Sheybanimail_outline

XXX Topic

Shashank Ballamail_outline

XXX Topic

Jung-Woo Changmail_outline

XXX Topic

Seira Hidanomail_outline

XXX Topic

Master

Benjamin Yangmail_outline

XXX Topic

Zhifei Xiemail_outline

XXX Topic

Jefferson Chienmail_outline

XXX Topic

Undergraduate

Position available

XXX Topic

Position available

XXX Topic

Position available

XXX Topic

Visiting Scholar

Seira Hidano

XXX Topic

Alumni

• Bita Rouhani, at Microsoft Research
• Mohammad Ghasemzadeh, at Apple
• Ebrahim M. Songhori (PhD), at Google
• Azalia Mirhoseini (PhD), now researcher at Google Brain
• Ye Zhang (MS), pursuing PhD at Rice University
• Chenhui Huang (MEE), now at Oracle
• Golsa Ghiaasi Hafezi (Post-doc), now Research Associate at Technische Universitat Wien.
• Joonho Kong (Post-doc), Assistant Professor at Kyungpook National University
• Arslan Munir (Post-doc), now Assistant Professor at the University of Nevada, Reno
• Masoud Rostami (PhD), now at Oracle Research
• Mehrdad Majzoobi (PhD), Founder and CEO of MeshMotion Inc.
• Salar Yazdjerdi (MEE), now at Uber Software Development
• Kangqiao Hu (MEE), now at AMD
• Ming Tong (MEE), now at C&J Energy Services
• Kai Li (MEE), now at Qualcomm
• Cathy Huang (BS), pursuing MBA at Duke
• Kevin Ting (BS), Customer Success Manager at HealthCrowd
• Praveen Kumar Pendyala (Undergrad Intern), pursuing MS at TU Darmstadt
• Akshat Kharaya (Undergrad Intern), now senior analyst at McKinsey & Company

Email:
farinaz@ucsd.edu

Address:
Electrical & Computer Engineering
University of California, San Diego
9500 Gilman Drive, MC 0407
Jacobs Hall, Room 6401
La Jolla, CA 92093-0407

Lab Location: EBU1-2514
University of California San Diego
9500 Gilman Dr, La Jolla, CA 92093