Training multimodal systems for classification with multiple objectives
Publication date
2020-01-01
Document type
Conference paper
Author
Organisational unit
Universität Bonn
Scopus ID
ISSN
Conference
1st International Workshop on Cross-lingual Event-centric Open Analytics co-located with the 17th Extended Semantic Web Conference (ESWC 2020) Heraklion, Crete, Greece, June 3, 2020
Series or journal
CEUR Workshop Proceedings
Periodical volume
2611
Book title
CLEOPATRA 2020: cross-lingual event-centric open analytics : proceedings of the 1st International Workshop on Cross-lingual Event-centric Open Analytics
First page
1
Last page
15
Peer-reviewed
✅
Part of the university bibliography
Nein
Keyword
Machine Learning
Multimodal Data
Probabilistic Method
Abstract
We learn about the world from a diverse range of sensory information. Automated systems lack this ability as investigation has centred on processing information presented in a single form. Adapting architectures to learn from multiple modalities creates the potential to learn rich representations of the world - but current multimodal systems only deliver marginal improvements on unimodal approaches. Neural networks learn sampling noise during training with the result that performance on unseen data is degraded. This research introduces a second objective over the multimodal fusion process learned with variational inference. Regularisation methods are implemented in the inner training loop to control variance and the modular structure stabilises performance as additional neurons are added to layers. This framework is evaluated on a multilabel classification task with textual and visual inputs to demonstrate the potential for multiple objectives and probabilistic methods to lower variance and improve generalisation.
Version
Not applicable (or unknown)
Access right on openHSU
Metadata only access