Incrementally learning new classes with generative classification

Información del proyecto

GenClassifier4CL

Identificador del acuerdo de subvención: 101067759

DOI

10.3030/101067759

Proyecto cerrado

Fecha de la firma de la CE 23 Junio 2022

Fecha de inicio 1 Octubre 2022

Fecha de finalización 30 Septiembre 2024

Financiado con arreglo a

• Marie Skłodowska-Curie Actions (MSCA)

Coste total

Sin datos

Aportación de la UE

€ 191 760,00

Coordinado por

KATHOLIEKE UNIVERSITEIT LEUVEN
Belgium

Objetivo

"Learning continually from non-stationary streams of data is a key feature of natural intelligence, but an unsolved problem in deep learning. Particularly challenging for deep neural networks is the problem of ""class-incremental learning"", whereby a network must learn to distinguish classes that are not observed together. In deep learning, the default approach to classification is learning discriminative classifiers. This works great in the i.i.d. setting when all classes are available simultaneously, but when new classes must be learned incrementally, successful training of discriminative classifiers depends on workarounds such as storing data or generative replay. In a radical shift of gears, here I propose to instead address class-incremental learning with generative classification. Key advantage is that generative classifiers – unlike discriminative classifiers – do not compare classes during training, but only during inference (i.e. when making a classification decision). As a proof-of-concept, in preliminary work I showed that a naïve implementation of a generative classifier, with a separate variational autoencoder model per class and likelihood estimation through importance sampling, outperforms comparable generative replay methods. To improve the efficiency, scalability, and performance of this generative classifier, I propose four further modifications: (1) move the generative modelling objective from the raw inputs to an intermediate network layer; (2) share the encoder network between classes, but not necessarily the decoder networks; (3) use fewer importance samples for unlikely classes; and (4) make classification decisions hierarchical. This way, during my MSCA fellowship hosted in the group of Prof Tinne Tuytelaars, I hope to develop generative classification into a practical, efficient, and scalable state-of-the-art deep learning method for class-incremental learning."

Ámbito científico (EuroSciVoc)

• ciencias naturales › informática y ciencias de la información › inteligencia artificial › aprendizaje automático › aprendizaje profundo

Palabras clave

• Continual learning; deep neural networks; catastrophic forgetting; generative classification; class-incremental learning

Programa(s)

• HORIZON.1.2 - Marie Skłodowska-Curie Actions (MSCA) PROGRAMA PRINCIPAL
  

Tema(s)

• HORIZON-MSCA-2021-PF-01-01 - MSCA Postdoctoral Fellowships 2021
  

Régimen de financiación

HORIZON-TMA-MSCA-PF-EF - HORIZON TMA MSCA Postdoctoral Fellowships - European Fellowships

Convocatoria de propuestas

HORIZON-MSCA-2021-PF-01

Coordinador

KATHOLIEKE UNIVERSITEIT LEUVEN

• Aportación neta de la UEn: € 191 760,00
• Dirección: OUDE MARKT 13 ; 3000 LEUVEN ; Bélgica
• Región: Vlaams Gewest › Prov. Vlaams-Brabant › Arr. Leuven
• Tipo de actividad: Higher or Secondary Education Establishments
• Coste total: Sin datos