The features of the architecture that will execute the inference are taken into account during the whole development process, starting from the early stages such as pre-training hyperparameter optimization and algorithm configuration.
The tool flow implements support for agile development methodologies, to be easily adopted by SMEs and midcaps.
The development process considers that the system should adapt to different operating modes at runtime.
The development process is conceived to support novel processing platforms to be exploitable beyond the end of the project.
The development process automates the introduction of algorithm features and programming techniques improving the resilience of the system to attacks.
In DL, good precision levels can be also obtained using algorithms with reduced complexity. All the optimization utilities in the ALOHA tool flow will consider the effects of algorithm simplification on precision, execution time, energy and power.
The ALOHA consortium organizes its first workshop dedicated to topics covered in the project development, such as Deep learning, Embedded systems, Security of Deep learning architectures. Speakers from Italy, Switzerland and Austria will share their expert knowledge and provide information about ALOHA tool development. The event is organized and hosted by SCCH - Software Competence Center Hagenberg GmbH and will take place in Linz, Austria on February 7th, 2019.
VISION is the world’s leading trade fair for machine vision. The who’s who in the machine vision industry meets in Stuttgart and presents an incomparable range of products and services: from sensors to processors, from cables to cameras, from software to illumination systems. However, complete machine vision systems and very specific applications for many different industries are exhibited – from mechanical engineering and the automotive industry through to the medical technology industry and many, many more industries besides.
During VISION 2018.(Messe Stuttgart, Germany, November 6-8), IridaLabs presented the ALOHA architecture and the project's first research findings.
Wild Patterns is a Tutorial on Adversarial Machine Learning, periodically held by Pluribus One during several top conferences and events on machine learning and computer vision. The secure machine-learning algorithms and the related security assessment procedures defined within the ALOHA project have been presented by Battista Biggio and Fabio Roli during several editions of the tutorial, within these events: Winter School on Quantitative Systems Biology: Learning and Artificial Intelligence (Trieste, Italy, November 15-16); ACM CCS 2018, 25th ACM Conference on Computer and Communications Security (Toronto, Canada, October 15-19); ECCV 2018, European Conference on Computer Vision (Munich, Germany, September 8-14); EUSIPCO 2018, 26th European Signal Processing Conference (Rome, Italy, September 3-7). Slides and videos of the editions of the Tutorial are available in this article.