Even Faster CNNs: Exploring the New Class of Winograd Algorithms

Tuesday, May 22, 1:30 PM - 2:30 PM
Summit Track: 
Technical Insights I
Mission City Ballroom B2-B5

Over the past decade, deep learning networks have revolutionized the task of classification and recognition in a broad area of applications. Deeper and more accurate networks have been proposed every year and more recent developments have shown how these workloads can be implemented on modern low-power embedded platforms. This presentation will discuss a recently introduced class of algorithms to reduce the arithmetic complexity of convolution layers with small filter sizes. After an introduction to the latest optimizations techniques for the most common solutions such as GEMM, the talk will dive deeply into the design of Winograd algorithms, analyzing the complexity and the performance achieved for convolutional neural networks.


Gian Marco Iodice

Senior Software Engineer of the Machine Learning Group, ARM

Gian Marco Iodice is a Senior Machine Learning Software Engineer at ARM on the Compute Library team. He received the MSc degree, with Honours, in Electronic Engineering from the University of Pisa (Italy) and has 3+ years of experience on researching and optimizing Machine Learning and Computer Vision algorithms on Arm CPUs and GPUs. In the last few years, Gian Marco has been a speaker for various conferences in USA and in Asia at the Embedded Vision Summit, ARM TechCon and ARM Tech Symposia where he discussed how the revolutionary and powerful Convolutional Neural Networks can be designed and optimized on ARM-based platforms.

See you at the Summit! May 20-23 in Santa Clara, California!
Register today and reserve your hotel room!