Modality-specific Learning Rates for Effective Multimodal Additive Late-fusion
Yiqun Yao, Rada Mihalcea
2022Findings of the Association for Computational Linguistics: ACL 202229 citationsDOIOpen Access PDF
Abstract
In multimodal machine learning, additive latefusion is a straightforward approach to combine the feature representations from different modalities, in which the final prediction can be formulated as the sum of unimodal predictions.
Topics & Concepts
ModalitiesModality (human–computer interaction)Computer scienceArtificial intelligenceFusionMachine learningFeature (linguistics)Multimodal learningLinguisticsSocial scienceSociologyPhilosophyMultimodal Machine Learning ApplicationsDomain Adaptation and Few-Shot LearningTopic Modeling