PODCAST – Garbage In, Garbage Out – High-Quality Datasets for Edge ML Research
https://www.buzzsprout.com/2363070/episodes/16955037
Datasets & Benchmarks
Our goals are to: Curate realistic, appropriately sized datasets—expanded through community collaboration Support open research into performance trade-offs (power, memory, accuracy, etc.) Foster shared learning and optimization across the ecosystem A public repository of datasets specific to edge AI and tinyML use cases will empower researchers, developers, and companies to evaluate their models with confidence. It will also help align innovations with real-world deployment scenarios. Importantly, the Working Group does not judge submission quality or run official benchmarks. Instead, our role is to enable honest, community-driven comparison—creating the infrastructure for innovation.
Members
Prof. Eiman Kanjo
professor in TinyML
Nottingham Trent Univeristy/ Imperial College London
e.kanjo@imperial.ac.uk
Rosina Haberl
Strategic Partner Senior Program Manager
EDGE AI FOUNDATION
rosina@edgeaifoundation.org
Meeting Minutes
Output
Leveling the Playing Field for Edge AI Research Through High Quality Datasets
Leveling the Playing Field for Edge AI Research Through High Quality Datasets Published by EDGE AI FOUNDATION Datasets & Benchmarks Working Group: Adam Fuks - NXP, Chair Petrut Bogdan - Innatera Vijay Janappa Reddi - Harvard University Eiman Kanjo - Imperial College Colby Banburry - Harvard University Sam Al Attiyah
