François Chung, Ph.D.

Tag: image classification

AWS: foundations and machine learning

AWS: foundations and machine learning

AWS training, MOOC (2020). These online courses provide an overall understanding of AWS cloud, with an overview of cloud concepts, services, security, architecture, pricing and support. Specific courses to AWS partners teach best practices to address customer business priorities around costs, agility, compliance, innovation and growth. Machine Learning (ML) is also covered, with a deep dive into the same ML curriculum used to train Amazon’s developers and data scientists.

AWS Cloud Practitioner Essentials

Main topics:

  • AWS core services;
  • AWS integrated services;
  • AWS architecture;
  • AWS security;
  • Pricing and support.

AWS Partner Solutions: Business foundations

Main topics:

  • Build your business with AWS;
  • What matters to AWS customers;
  • Security, identity and compliance;
  • Pricing and licensing;
  • Migration and cloud adoption;
  • Opportunity management.

AWS Partner Solutions: Technical foundations

Main topics:

  • AWS solution architects;
  • AWS architectural concepts;
  • Building blocks;
  • AWS Well-Architected Framework;
  • Architecting an AWS solution;
  • Engaging customers and architecting solutions.

AWS Machine Learning: Decision maker

Main topics:

  • Demystifying AI/ML/DL;
  • ML for business challenges;
  • ML terminology;
  • Exploring the ML toolset.

AWS Machine Learning: Data scientist

Main topics:

  • Math for ML;
  • Linear and logistic regression;
  • Elements of data science;
  • Real world ML decisions.

References

EUE 2017 - Book

EUE 2017 – Book

Publication

François Chung, Tomás Rodríguez; Multi-focal Image Segmentation, Classification and Authentication: A General Framework applied on Microscope Pollen Images; Éditions universitaires européennes (EUE), Saarbrücken, 2017; ISBN: 978-3841677907.

Abstract

In this book, we propose a general framework for multi-focal image segmentation, classification and authentication, the methodology being demonstrated on microscope pollen images. The framework is meant to be generic and based on a brute force-like approach aimed to be efficient not only on any kind, and any number, of pollen images (regardless of the pollen type), but also on any kind of multi-focal images. All stages of the framework are designed to be used in an automatic fashion. First, the optimal focus is selected using the absolute gradient method. Then, pollen grains are extracted from microscope images, followed by the automatic segmentation of their exine. A coarse-to-fine approach ensures a smooth and accurate segmentation of both structures. Finally, feature extraction and selection are performed on pollen grains using a generalized approach and the pollen classification is tested with four classifiers: Weighted Neighbor Distance, Neural Network, Decision Tree and Random Forest. The latter method, which has shown the best and more robust classification accuracy results (above 97% for any number of pollen types), is finally used for a final authentication stage.

References

Publication

Related article

APIFRESH (Inspiralia project)

Learn more

EUE – Éditions universitaires européennes

APIFRESH

APIFRESH

Inspiralia project @Madrid, Spain (2013). The competitiveness of the European beekeeping sector is progressively falling due to the reduction of production as a direct consequence of the decrease in bee population. In addition, beekeeping products from countries with lower quality standards are gaining market share in Europe through an unfair competition. Furthermore, there is a lack of standards at European level for certain bee products like pollen and royal jelly.

This means that it is possible to find in the market products under these labels without any control of quality and authenticity. Few countries in Europe have some guidelines or regional standards for products other than honey, which results in a lack of standardization at the European level.

Therefore, the objectives of the APIFRESH project are threefold:

  • to develop European standards for bee pollen and royal jelly;
  • to establish health-relevant criteria for pollen and royal jelly;
  • to determine the authenticity of both pollen and honey.

Partners of this project include:

  • Balparmak (TR);
  • Campomiel (ES);
  • Centro Agrario de Marchamalo (ES);
  • CTC - Centro Tecnológico Nacional de la Conserva y Alimentación (ES);
  • EPBA - European Professional Beekeepers Association (EU);
  • FNAP - Federação Nacional dos Apicultores de Portugal (PT);
  • Inspiralia (ES);
  • OMME - Országos Magyar Méhészeti Egyesület (HU);
  • Parco Tecnologico Padano (IT);
  • TÜBITAK-MAM - Türkiye Bilimsel ve Teknolojik Araştırma Kurumu-Marmara Araştırma Merkezi (TR).

In this project, my work consists in the development of a software for the bee pollen classification and authentication. In a first step, bee pollen loads captured from a camera are separated by pollen type using a classification based on color. In a second step, a microscope is used to capture an accurate image of pollen grains from which discriminative features are extracted to identify the pollen origin, i.e. by considering the pollen grain as belonging either to a known type (pollen classification) or to an unknown type (outlier detection).

References

Related articles

EUE 2017 (book)
COMPAG 2015 (academic journal article)
Micron 2015 (academic journal article)
Inspiralia 2013 (technical report)
Inspiralia 2012 (technical report)

Learn more

Inspiralia 2013 - Technical report

Inspiralia 2013 – Technical report

Publication

François Chung, Tomás Rodríguez; A general framework for multi-focal image classification and authentication: Application to microscope pollen images; Inspiralia, Madrid, 2013.

Abstract

In this article, we propose a general framework for multi-focal image classification and authentication, the methodology being demonstrated on microscope pollen images. The framework is meant to be generic and based on a brute force-like approach aimed to be efficient not only on any kind, and any number, of pollen images (regardless of the pollen type), but also on any kind of multi-focal images. All stages of the framework's pipeline are designed to be used in an automatic fashion. First, the optimal focus is selected using the absolute gradient method. Then, pollen grains are extracted using a coarse-to-fine approach involving both clustering and morphological techniques (coarse stage), and a snake-based segmentation (fine stage). Finally, features are extracted and selected using a generalized approach, and their classification is tested with four classifiers: Weighted Neighbor Distance, Neural Network, Decision Tree and Random Forest. The latter method, which has shown the best and more robust classification accuracy results (above 97% for any number of pollen types), is finally used for the authentication stage.

Keywords

  • generalized feature extraction
  • image classification
  • microscope images
  • optimal focus selection
  • pollen authentication
  • Random Forest
  • snake-based segmentation
  • supervised clustering

References

Publication

Related articles

APIFRESH (Inspiralia project)
EUE 2017 (book)
Inspiralia 2012 (technical report)

Learn more