About Me

Byung Hoon Ahn

I am a Ph.D. Candidate in Computer Science at UCSD working with Prof. Hadi Esmaeilzadeh. During my Ph.D., I have published multiple papers in the field of Machine Learning and Computer Architecture: ICLR, MLSys, and MICRO. I was selected Finalist of Qualcomm Innovation Fellowship in 2020 and recipient of Korean Government Scholarship for Overseas Studies in 2018.

I have 5+ years of industry experience as an engineer in various layers of the systems stack for AI/ML. I have worked on optimizing deep learning models (quantization, pruning, and graph optimization). I have hands-on experience developing and optimizing the innards of LLVM and TVM. I also have profound experience developing models using TensorFlow and PyTorch, and deploying models using ONNX and TFLite.

I have successfully completed multiple internships at Qualcomm AI Research, and I am currently working as a Research Scientist at Protopia AI, an AI privacy start-up. Prior to my Ph.D. studies, I got my B.Eng in Electical Engineering from Korea University, and worked at Samsung Research as a Software Engineer.

Education

Ph.D. in Computer Science. University of California, San Diego
Sep. 2018 – present
Area: Computer Architecture, Compiler Optimization, and Machine Learning
Advisor: Prof. Hadi Esmaeilzadeh
Topic: Abstractions and Mechanisms for AI-Enabled Compilation of Deep Neural Networks [thesis proposal]

M.S. in Computer Science. University of California, San Diego
Sep. 2018 – Jun. 2020
Area: Computer Architecture, Compiler Optimization, and Machine Learning
Advisor: Prof. Hadi Esmaeilzadeh
Topic: Compilation and Optimization Techniques for Machine Learning Workloads [research exam]

B.Eng. in Electrical Engineering. Korea University
Mar. 2009 – Feb. 2015
Area: Computer Vision and Computer Architecture
Advisor: Prof. Sung-Jea Ko


Honors and Awards

  • Qualcomm Innovation Fellowship Finalist. Qualcomm (2020)
  • Korean Government Scholarship for Overseas Studies. NIIED, Ministry of Education (2018-2020)
  • CSE Department Fellowship, UC San Diego (2018)
  • Super Rookie Award. Samsung Research (2015)
  • Teaching Service Scholarship. Korea University (2014)
  • Semester High Honors. Korea University (3 semesters, 2011-2014)
  • Honors Scholarship. Korea University (4 semesters, 2009-2014)

Publications

  • Protopia AI: Taking on the Missing Link in AI Privacy and Data Protection
    Byung Hoon Ahn, DoangJoo Synn, Masih Derkani, Eiman Ebrahimi, Hadi Esmaeilzadeh
    NeurIPS Demonstrations, 2021 [link,colab,demo]
  • FastStereoNet: A Fast Neural Architecture Search for Improving the Inference of Disparity Estimation on Resource-Limited Platforms
    Mohammad Loni, Ali Zoljodi, Amin Majd, Byung Hoon Ahn, Masoud Daneshtalab, Mikael Sjödin, Hadi Esmaeilzadeh
    IEEE SMC, IF=13.45, 2021 [paper]
  • Planaria: Dynamic Architecture Fission for Spatial Multi-Tenant Acceleration of Deep Neural Networks
    Soroush Ghodrati, Byung Hoon Ahn, Joon Kyung Kim, Sean Kinzer, Brahmendra Yatham, Navateja Alla, Hardik Sharma, Mohammad Alian, Eiman Ebrahimi, Nam Sung Kim, Cliff Young, Hadi Esmaeilzadeh
    MICRO, 2020 [paper,video]
  • Chameleon: Adaptive Code Optimization for Expedited Deep Neural Network Compilation
    Byung Hoon Ahn, Prannoy Pilligundla, Amir Yazdanbakhsh, Hadi Esmaeilzadeh
    ICLR, 2020 [paper,video]
  • Ordering Chaos: Memory-Aware Scheduling of Irregularly Wired Neural Networks for Edge Devices
    Byung Hoon Ahn, Jinwon Lee, Jamie Menjay Lin, Hsin-Pai Cheng, Jilei Hou, Hadi Esmaeilzadeh
    MLSys, 2020 [paper,slide]
    * Featured in Qualcomm’s Invited Talk in EMC² Workshop at NeurIPS 2019 [link,video]
  • Reinforcement Learning and Adaptive Sampling for Optimized DNN Compilation
    Byung Hoon Ahn, Prannoy Pilligundla, Hadi Esmaeilzadeh
    ICML Workshop on Reinforcement Learning for Real Life, 2019 [paper]

Work/Research Experience

Graduate Student Researcher. Alternative Computing Technologies (ACT) Lab.
Aug. 2018 – present (3 yrs 4 mos)
Advisor: Prof. Hadi Esmaeilzadeh

Research Scientist. Protopia AI
Oct. 2020 – present (1 yr 3 mos)
CEO: Dr. Eiman Ebrahimi

Research Intern. Qualcomm AI Research
Jun. 2020 – Sep. 2020 (3 mos)
Mentor: Dr. Harris Teague. Manager: Dr. Chris Lott, Dr. Jilei Hou

Research Intern. Qualcomm AI Research
Jul. 2019 – Sep. 2019 (3 mos)
Mentor: Dr. Jinwon Lee. Manager: Dr. Chris Lott, Dr. Jilei Hou

Software Engineer. Samsung Research
Jan. 2015 – Jun. 2018 (3 yrs 6 mos)
Mentor: Dr. Yeonbok Lee. Manager: Dr. Myungsun Kim

Research Intern. Computer Vision & Image Processing (CVIP) Lab.
Jul. 2014 – Dec. 2014 (6 mos)
Advisor: Prof. Sung-Jea Ko


Military Experience

Sergeant (E-5). 2nd Combat Aviation Brigade. 2nd ID. 8th US Army
Jul. 2011 – Apr. 2013 (21 mos)


Teaching Experience

Teaching Assistant
CSE141 Introduction to Computer Architecture, UC San Diego (Spring 2020)
KECE382 Control Engineering, Korea University (Fall 2014)
KECE231 Engineering Mathematics, Korea University (Spring 2014)

Graduate Student Mentor
Early Research Scholars Program (ERSP), UC San Diego (2018-2019)

Part-time Teacher
Mathematics, Jangchung Middle School with Samsung Dreamclass (2014)