VIS4ML Research

VIS4ML research explores the intersection of visualization and machine learning, aiming to improve the understanding, development, and deployment of ML models. Current efforts focus on using visualizations to interpret model behavior, particularly concerning data quality and model performance across various data types, and on optimizing ML model implementation on FPGAs for low-latency applications using architectures like transformers and recurrent neural networks. This research is significant for enhancing the transparency and efficiency of ML, with applications ranging from high-energy physics and autonomous vehicles to scientific computing.

Papers

September 8, 2024

Low Latency Transformer Inference on FPGAs for Physics Applications with hls4ml
Zhixing Jiang, Dennis Yin, Yihui Chen, Elham E Khoda, Scott Hauck, Shih-Chieh Hsu, Ekaterina Govorkova, Philip Harris, Vladimir Loncar, Eric A. Moreno
Transformer Architecture Field Programmable Gate Array Transformer Inference VIS4ML Research

September 4, 2024

Adversarial Attacks on Machine Learning-Aided Visualizations
Takanori Fujiwara, Kostiantyn Kucher, Junpeng Wang, Rafael M. Martins, Andreas Kerren, Anders Ynnerman
Adversarial Attack New Machine Relevant Visualization Adversarial Machine Learning VIS4ML Research

August 9, 2024

rule4ml: An Open-Source Tool for Resource Utilization and Latency Estimation for ML Models on FPGA
Mohammad Mehdi Rahimifar, Hamza Ezzaoui Rahali, Audrey C. Therrien
Machine Learning Model Field Programmable Gate Array Neural Network Architecture Inference Latency Resource Utilization High Level Synthesis VIS4ML Research

August 10, 2023

Are We Closing the Loop Yet? Gaps in the Generalizability of VIS4ML Research
Hariharan Subramonyam, Jessica Hullman
Stronger Generalizability Human in the Loop Filling Gap VIS4ML Research

July 15, 2023

Visual Analytics For Machine Learning: A Data Perspective Survey
Junpeng Wang, Shixia Liu, Wei Zhang
Machine Learning Machine Learning Model Machine Learning Pipeline Visual Analytics Data Perspective VIS4ML Research

June 1, 2023

CS4ML: A general framework for active learning with arbitrary data based on Christoffel functions
Ben Adcock, Juan M. Cardenas, Nick Dexter
Active Learning General Framework Adaptive Sampling VIS4ML Research Approximation Space Gradient Amplification

July 1, 2022

Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml
Elham E Khoda, Dylan Rankin, Rafael Teixeira de Lima, Philip Harris, Scott Hauck, Shih-Chieh Hsu, Michael Kagan, Vladimir Loncar, Chaitanya Paikara, Richa Rao, Sioni Summers, Caterina Vernieri, Aaron Wang
Recurrent Neural Network Long Short Term Memory Field Programmable Gate Array Low Latency Neural Network Inference Recurrent Neural Network Architecture Recurrent Layer Recurrent Unit VIS4ML Research

May 16, 2022

Real-time semantic segmentation on FPGAs for autonomous vehicles with hls4ml
Nicolò Ghielmetti, Vladimir Loncar, Maurizio Pierini, Marcel Roed, Sioni Summers, Thea Aarrestad, Christoffer Petersson, Hampus Linander, Jennifer Ngadiuba, Kelvin Lin, Philip Harris
Semantic Segmentation Autonomous Driving Autonomous Vehicle Real Time Field Programmable Gate Array Convolutional Layer Convolutional Neural Network Architecture VIS4ML Research