Close

Session

Research Manuscript
: Faster, Safer, Greener: AI-driven Evolution in Smart Edge
Session Chairs
Peipei Zhou
Brown University
Pi-Cheng Hsiu
Academia Sinica
DescriptionThis session explores AI-driven innovations shaping edge intelligence in acceleration, reliability, and sustainability. It begins with acceleration in 3D point cloud detection, edge-compute-free image compression, and multi-model decentralized federated learning. Reliability is then reinforced with Linux runtime integrity measurement, automated hardware verification, and high-level synthesis with post-implementation. It concludes with sustainability through energy-efficient forecasting and architecture-level power modeling. These innovations drive the next generation of smart edge computing.
Event Type
Research Manuscript
TimeWednesday, June 253:30pm - 5:30pm PDT
Location3006, Level 3
Topics
Systems
Tracks
SYS4: Embedded System Design Tools and Methodologies
Similar Sessions
From Simulation to Threats, a Deep-dive into CPS and IoT Design
Smarter Compute, Faster Inference: Optimizing AI Systems on Edge
AI Under Attack: Enhancing Privacy, Robustness, and Trust in ML Systems
Presentations
3:30pm - 3:45pm PDTDAWN: Accelerating Point Cloud Object Detection via Object-Aware Partitioning and 3D Similarity-Based Filtering
Authors
Dongdong Tang
Yu Mao
Weilan Wang
Nan Guan
Tei-Wei Kuo
Chun Jason Xue
3:45pm - 4:00pm PDTEasz: An Agile Transformer-based Image Compression Framework for Resource-constrained IoTs
Authors
Yu Mao
Jingzong Li
Jun Wang
Hong Xu
Tei-Wei Kuo
Guan Nan
Chun Xue
4:00pm - 4:15pm PDTMMDFL: Multi-Model-based Decentralized Federated Learning for Resource-Constrained AIoT Systems
Authors
DengKe Yan
Yanxin Yang
Ming Hu
Xin Fu
Mingsong Chen
4:15pm - 4:30pm PDTLightRIM: Light Runtime Integrity Measurement for Linux Kernels in Embedded Applications
Authors
Yili Guo
Zhuoran Ma
Xiangyue Li
Jiajia Huang
Wanli Chang
4:30pm - 4:45pm PDTUVLLM: An Automated Universal RTL Verification Framework using LLMs
Authors
Yuchen Hu
Junhao Ye
Ke Xu
Jialin Sun
Shiyue Zhang
Xinyao Jiao
Dingrong Pan
Jie Zhou
Ning Wang
Weiwei Shan
Xinwei Fang
Xi Wang
Nan Guan
Zhe Jiang
4:45pm - 5:00pm PDTA Post-Implementation Performance Prediction Method with HLS Optimization Directives
Authors
Jingyu Zhu
Yan Ding
Lu Xiao
Kenli Li
Chubo Liu
Zheng Xiao
5:00pm - 5:15pm PDTContention-Aware Forecasting of Energy Efficiency through Sequence-Based Models in Modern Heterogeneous Processors
Authors
Mohammed Bakr Sikal
Jeferson Gonzalez
Heba Khdr
Joerg Henkel
5:15pm - 5:30pm PDTAutoPower: Automated Few-Shot Architecture-Level Power Modeling by Power Group Decoupling
Authors
Qijun Zhang
Yao Lu
Mengming Li
Zhiyao Xie