News
Profs. Gross and Meyer give deep learning tutorial at IEEE EPEPS 2019
Profs. Warren Gross and Brett H. Meyer presented a tutorial on the optimization of hardare and software for deep learning at IEEE EPEPS 2019 in Montreal today. Gross introduced machine learning in general, and deep learning in particular, from a computational perspective. He then summarized recent work on custom architecture for DNN acceleration. Meyer followed up with an introduction to multi-objective hyperparameter optimization, with a focus on deployment to low-cost IoT processors.
Prof. Meyer presents at Dawson Humanities and Public Life Conference
Prof. Brett H. Meyer presented at the Dawson College Humanities and Public Life Conference today, giving a talk entitled ‘The Algorithms Aren’t Alright: Why Machine Learning Still Needs Us.’ In it he introduced machine learning in general, deep learning in particular, and highlighted some of the challenges that arise in the application of deep learning: robustness, explainaibility, and bias.
Poster: Optimizing Keyword Spotting on Microcontrollers
Undergraduate Adithya Lakshminarayanan presented his summer research at the SURE Poster Session today. It explores the different trade-offs inherent in quantizing models for keyword spotting. FP is best for top accuracy, but 8- and 16-bit models each strike interesting trade-offs under tighter constraints.
New website!
Please bear with us as we move our website to Jekyll and slowly add content.