Cheng Gong       

Cheng Gong

Assistant Researcher & Postdoctoral Fellow
College of Software, Nankai University, Tianjin, China
E-mail:  cheng-gong@nankai.edu.cn
Address: 38 Tongyan Road, Jinnan District, Tianjin, China 300350
Phone: +86 156-2062-5044
Github | GoogleScholar | ResearchGate | DBLP | ORCID | SemanticScholar | 中文主页

Research Interests

  • Neural network compression
  • Efficient neural network inference
  • Heterogeneous computing

Awards and Commendables

  1. 2023 Excellent Doctoral Dissertation Award, Tianjin Chapter, ACM China
  2. 2021 Second Prize Scholarship, Nankai University
  3. 2020 Second Prize Scholarship, Nankai University
  4. 2019 First Prize Scholarship, Nankai University
  5. 2015 National Encouragement Scholarship for Undergraduate Students, Nankai University
  6. 2014 Hezhan Incentive Scholarship for Undergraduate Students, Nankai University

Education

  • PhD (2016 - 2022) in Nankai University, Tianjin, China
  • BS (2012 - 2016) in Nankai University, Tianjin, China

Employments

  • College of Software, Nankai University, Tianjin, China
    • Assistant Researcher & Postdoctoral Fellow (July 2022 - Present)

Publications

  1. [ECCV'24] Cheng Gong, Yao Chen, Qiuyang Luo, Ye Lu, Tao Li, Yuzhi Zhang, Yufei Sun, and Le Zhang. Deep Feature Surgery: Towards Accurate and Efficient Multi-Exit Networks. [Code][Paper]
  2. [CAAI AIR'24] Cheng Gong, Haoshuai Zheng, Mengting Hu, Zheng Lin, Deng-Ping Fan, Yuzhi Zhang, and Tao Li. Minimize Quantization Output Error with Bias Compensation. (Accepted) [Code][Paper]
  3. [JCST'24] Cheng Gong, Ye Lu, Su-Rong Dai, Qian Deng, Cheng-Kun Du, and Tao Li. AutoQNN: An End-to-End Framework for Automatically Quantizing Neural Networks. [Code] [Paper]
  4. [TPDS'22] Cheng Gong, Ye Lu, Kunpeng Xie, Zongming Jin, Tao Li, and Yanzhi Wang. Elastic Significant Bit Quantization and Acceleration for Deep Neural Networks. [Code] [Paper]
  5. [TC'21] Cheng Gong, Yao Chen, Ye Lu, Tao Li, Cong Hao, and Deming Chen. VecQ: Minimal Loss DNN Model Compression with Vectorized Weight Quantization. [Code] [Paper]
  6. [Journal of Software'21] Cheng Gong, Ye Lu, Surong Dai, Fangxin Liu, Xinwei Chen, and Tao Li. Ultra-Low Loss Quantization Method for Deep Neural Network Compression. [Code] [Paper]
  7. [Information Science'21] Cheng Gong, Ye Lu, Chunying Song, Tao Li, and Kai Wang. OSN: Onion-Ring Support Neighbors for Correspondence Selection. [Code] [Paper]
  8. [Communications of the CCF'21] Ye Lu, Cheng Gong, and Tao Li. Challenges and Opportunities in Automatic Compression of Deep Neural Networks. [Paper]
  9. [IJCNN'19] Cheng Gong, Tao Li, Ye Lu, Cong Hao, Xiaofan Zhang, Deming Chen, and Yao Chen. $\mu$L2Q: An Ultra-Low Loss Quantization Method for DNN Compression. [Code] [Paper]
  10. [ISVLSI'19] Yao Chen, Kai Zhang, Cheng Gong, Cong Hao, Xiaofan Zhang, Tao Li, and Deming Chen. T-DLA: An Open-Source Deep Learning Accelerator for Ternarized DNN Models on Embedded FPGA. [Code] [Paper]
  11. [ICPADS'19] Fangxin Liu, Kunpeng Xie, Cheng Gong, Shusheng Liu, Ye Lu, and Tao Li. LHC: A Low-Power Heterogeneous Computing Method on Neural Network Accelerator. [Paper]
  12. [IJCAI'18] Deng-Ping Fan, Cheng Gong, Yang Cao, Bo Ren, Ming-Ming Cheng, and Ali Borji. Enhanced-alignment Measure for Binary Foreground Map Evaluation. [Code] [Paper]

Invited Talks

  1. Quantization Methods for Deep Neural Networks. CCF Chips Conference - Session of Outstanding Young Scholars in Computer Architecture, Nanjing, China, 2022.
  2. An Ultra Low-Loss Deep Neural Network Quantization Method. The 18th China Computer Federation Embedded System Technology Conference, CCF ESTC 2020, Chengdu, China, 2020.
  3. μl2Q: An Ultra Low Loss Quantization Method for DNN Compression. The 2019 International Joint Conference on Neural Networks, IJCNN. Budapest, Hungary, 2019.

Patents

  1. Multi-Classifier Neural NetworkOptimization via Feature Rerouting. Cheng Gong, Tao Li, and Ye L. patent No. CN202210507772.9.
  2. Elastic Significant Bit Quantiza-tion and Acceleration for Deep Neural Networks. Cheng Gong, Ye Lu, and Tao Li. patent No. ZL202010661226.1.

Major Projects & Funding

As Principle Investigator (Pl)

  1. Adaptive Nonlinear Quantization and Acceleration for Deep Neural Networks.
    China Postdoctoral Science Foundation. RMB ¥80,000: 2023 - 2025.

As Collaborator

  1. High-Performance Virtual Execution Engine for Smart Contracts and Trustworthy Heterogeneous Acceleration.
    National Natural Science Foundation of China. 2024 - 2027.
  2. Automatic DNN Quantization and Deployment for Domestic Intelligent Processors.
    Open Project Fund of State Key Laboratory of Computer Architecture, ICT, CAS. 2020 - 2022.
  3. Small Object Detection based on Deep Learning and Its Heterogeneous Acceleration.
    National Natural Science Foundation of China. 2019 - 2022.
  4. DNN Compression, Deployment, and Acceleration on Edge Devices and Cambrian Neural Processor.
    Open Project Fund of State Key Laboratory of Computer Architecture, ICT, CAS. 2019 - 2021.
  5. DNN Model Compression and Acceleration for Edge Intelligence.
    Tianjin Natural Science Youth Fund. 2019 - 2021.
  6. Heterogeneous Internet of Things Application Service Collaborative Technology for Deep Integration.
    National Key Research and Development Program of China. 2019 - 2023.
  • High-Level Language C++ Programming for Undergraduate Students. College of Software, Nankai University, Tianjin, China.
  • Deep Learning for Graduate Students and International Students. College of Software, Nankai University, Tianjin, China.

Professional Services

  • Conference Review
    • Qeios 2025
    • ACMMM Workshop UAVs in Multimedia (UAVM) 2024
    • Chinese Conference on Pattern Recognition and Computer Vision (PRCV) 2024

News

  • [07/2024] Our paper entitled "Deep Feature Surgery: Towards Accurate and Efficient Multi-Exit Networks" has been accepted by ECCV 2024.
  • [06/2024] Our paper entitled "Minimize Quantization Output Error with Bias Compensation" has been accepted by CAAI AIR 2024.