Chengcheng Li

Availability: I can commence employment immediately upon your decision.

Email: adz9xh@r.postjobfree.com, Phone: 865-***-****

Google Scholar: https://scholar.google.com/citations?hl=en&user=86i2dWoAAAAJ LinkedIn: https://www.linkedin.com/in/chengcheng-l-058b3a109/ EDUCATION

Ph.D. in Computer Engineering (GPA: 3.92) Fall 2015 { May 2023 University of Tennessee, Knoxville, Tennessee, USA WORK EXPERIENCE

Research Engineer Intern, TuSimple: Autonomous Trucking May - August 2022 Formulated ill-de ned problems with machine learning methods to solve challenging scenarios for self- driving truck business. Built benchmark systems to test the algorithms and models. Closely collaborated with other teams to conduct system design, integration, and tests. Graduate Research Assistant, University of Tennessee Fall 2015 - Present Developed algorithms and applications in the general area of computer vision and machine learning, with a speci c focus on e cient deep learning (network pruning, knowledge distillation). Graduate Teaching Assistant, University of Tennessee Fall 2016 - Fall 2018 Conducted holding o ce hours and grading assignments for ECE 315 Signals and Systems 1, ECE 316 Signals and Systems 2, and ECE 599/692 Deep Learning. SKILLS

Machine learning platforms (PyTorch, TensorFlow, and Keras) Programming languages (Python, C++)

Network structures (Di usion models, Transformer, CNN, RNN, and GAN) Tasks (Generative AI, images classi cation, anomaly detection, time series prediction Large-scale data analytics tool (Spark, Pandas)

SQL, cloud APIs (GCP, AWS), Git

PROJECTS AND CONTRIBUTIONS

Trajectory Prediction for Autonomous Driving

{ Build trajectory prediction models for self-driving trucks using machine learning methods. Ex- perienced data collection and processing, model training, performance evaluation, and error analysis. Closely collaborated with other teams to conduct system design, integration, and tests.

E cient Deep Learning

{ Network Pruning: 1) studied network pruning from the perspective of redundancy reduction and proposed a network pruning approach that identi es structural redundancy of a CNN and prunes lers in the selected layer(s) with the most redundancy. 2) developed a computationally e cient framework for structured pruning by exploring the potential of leveraging the intermediate results generated during the ne-tuning. [1,5]

{ Knowledge distillation: propose methods for online knowledge distillation by leveraging histor- ical information encoded in the training procedure. Trained a group of student networks from scratch in a peer-teaching manner. Conducted evaluation of the proposed methods with various structures and datasets. [2,7]

Generative Adversarial Networks (GAN)

{ Proposed and implemented a fast-converging conditional GAN (FC-GAN) to accelerate the convergence speed for conditional image synthesis. [3]

Event Detection and Recognition for Smart Grid

{ Processed, cleaned, and labeled raw frequency data collected from a power system. Developed a CNN-based detection and recognition algorithm that enables early, accurate, and robust detec- tion, recognition, and temporal localization of multi-type events in large-scale power systems.

[6]

Accelerated Dynamics of Atomistic Simulations

{ Implemented an RNN-based regression to accelerate the simulation clock of adaptive Ab initio molecular dynamics (a versatile and reliable computational approach to atomic-scale materials science). [4]

SELECTED PUBLICATIONS

1. Wang, Zi, Chengcheng Li, and Xiangyang Wang. \Convolutional neural network pruning with structural redundancy reduction." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021.

2. Li, Chengcheng, Zi Wang, and Hairong Qi. \Online Knowledge Distillation by Temporal-Spatial Boosting." Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 2022.

3. Li, Chengcheng, Zi Wang, and Hairong Qi. \Fast-converging conditional generative adversarial networks for image synthesis." In 2018 25th IEEE International Conference on Image Processing

(ICIP), pp. 2132-2136. IEEE, 2018.

4. Jiaqi Wang*, Chengcheng Li*, Seungha Shin, and Hairong Qi. \Accelerated Atomic Data Pro- duction in Ab Initio Molecular Dynamics with Recurrent Neural Network for Materials Research." The Journal of Physical Chemistry C 124.27 (2020): 148**-*****. *: equal contribution 5. Li, Chengcheng, Zi Wang, and Hairong Qi. \An E cient Pipeline for Pruning Convolutional Neural Networks." 20th IEEE International Conference on Machine Learning and Applications

(2020).

6. Li, Chengcheng, et al. Early Alarm: Robust Event Analysis for Power Systems using 1-D Fully Convolutional Network. Accepted by SmartGridComm 2023. 7. Li, Chengcheng, Zi Wang, Hairong Qi, 2022 International Joint Conference on Neural Networks HONORS AND AWARDS

Extraordinary Professional Promise, University of Tennessee, 2020 Min H. Kao Fellowship (6 out of 300), EECS Department, 2019-2021 Outstanding Graduate Teaching Assistants (3 out of 65), EECS Department, 2017 Meritorious Winner in Mathematical Contest in Modeling, 2011 SERVICE & ACTIVITIES

Served as reviewer for IEEE International Conference on Image Processing (ICIP),Winter Conference on Applications of Computer Vision (WACV), Association for the Advancement of Arti cial Intelligence

(AAAI)