I-Ting Lin's Website

About Me

I'm currently a graduate student researcher at Digital Circuits and Systems Lab at National Taiwan University (NTU), advised by Prof. Chia-Hsiang Yang. I received my B.S. in Electrical Engineering from NTU in June 2020, and I am pursuing my M.S. degree in Electronics Engineering also at NTU. My research interests are energy-efficient VLSI design and domain-specific computer architecture for emerging fields, including robotics, machine learning, and computer vision.

Publications

A 28nm 142mW Motion-Control SoC for Autonomous Mobile Robots

Authors: I-T. Lin, Z.-S. Fu, W.-C. Chen, L.-Y. Lin, N.-S. Chang, C.-P. Lin, C.-S. Chen, C.-H. Yang.
Accepted by International Solid-State Circuits Conference (ISSCC) 2023.

Details are coming soon!

A 28-nm 25.1TOPS/W Sparsity-Aware CNN-GCN Deep Learning SoC for Mobile Augmented Reality

Authors: W.-C. Huang, I-T. Lin, W.-C. Chen, L.-Y. Lin, N.-S. Chang, C.-P. Lin, C.-S. Chen, C.-H. Yang.
Published in International Symposium on VLSI Circuits (VLSI Circuits) 2022.

This work presents the first CNN-GCN SoC for diverse AI vision computations on mobile augmented reality (AR). The SoC outperforms prior CNN accelerators by 2x in energy efficiency and prior action recognition processor by 18x in latency.

Link to paper

Experience

Graduate Student Researcher,
Digital Circuits and Systems Lab

NTU | Sep. 2020 - present | Advisor: Prof. Chia-Hsiang Yang

I mainly work on domain-specific hardware acceleration and algorithm-hardware co-design.

Selected Projects:
- Motion-Control SoC for Autonomous Mobile Robots
- Deep Learnining SoC for Mobile Augmented Reality
- Hardware-Aware Neural Network Quantization

Software Application Engineering Intern, Ganzin Technology

New Taipei City, Taiwan | Aug. 2019

I was responsible to build a visual-searching testing system, incorporating Ganzin eye-tracking solution, for detecting and evaluating possible visual field loss

Undergraduate Student Researcher, Media IC & System Lab

NTU | Sep. 2018 - June 2019 | Advisor: Prof. Shao-Yi Chien

I worked on mobile deep learning and developed an automatic workflow to deploy CNNs on Android mobile phones

Selected Projects

Motion-Control SoC for Autonomous Mobile Robots

- Proposed a trajectory optimization accelerator
- Interated an ARM Cortex-M3 MCU
- Taped-out in 28nm CMOS technology
- Accepted by ISSCC'23

CNN-GCN Deep Learning SoC for Mobile Augmented Reality

- Integrated a CNN Engine, a GCN Engine, and a RISC-V MCU
- Taped-out in 28nm CMOS technology
- Published in VLSI'22

Hardware-Aware Neural Network Quantization

- Proposed a two-stage 4-bit quantization scheme
- Optimized quantized convolution for efficient HW design
- Honored by 2020 Delta Young Talent Award

Thermal Image Processing Module for Privacy-Preserving Human Detection

- Incorporated 4 ultra-low-resolution thermal sensors
- Optimized a thermal image SRCNN for ADP-Corvette-T1
- Currently the finalist of 2022 Andes Award

Feature-Based Image Morphing Accelerator on PYNQ FPGA

- Designed an accelerator for Beier-Nealy Algorithm
- Using Vivado High-Level Synthesis
- Ranked No.2 at the final project of Multimedia SoC Design

5-Stage Pipeline MIPS CPU

- Optimized the radix of Booth's Multiplier
- Conducted L2 Cache Analysis
- Final project of Digital System Design

Hello, My name is I-Ting Lin