Curriculum Learning for Compositional Visual Reasoning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Curriculum Learning for Compositional Visual Reasoning

Résumé

Visual Question Answering (VQA) is a complex task requiring large datasets and expensive training. Neural Module Networks (NMN) first translate the question to a reasoning path, then follow that path to analyze the image and provide an answer. We propose an NMN method that relies on predefined cross-modal embeddings to “warm start” learning on the GQA dataset, then focus on Curriculum Learning (CL) as a way to improve training and make a better use of the data. Several difficulty criteria are employed for defining CL methods. We show that by an appropriate selection of the CL method the cost of training and the amount of training data can be greatly reduced, with a limited impact on the final VQA accuracy. Furthermore, we introduce intermediate losses during training and find that this allows to simplify the CL strategy.
Fichier principal
Vignette du fichier
Curriculum_Learning_for_Compositional_Visual_Reasoning.pdf (306.39 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04045405 , version 1 (24-03-2023)

Identifiants

Citer

Wafa Aissa, Marin Ferecatu, Michel Crucianu. Curriculum Learning for Compositional Visual Reasoning. Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISAPP, Feb 2023, Lisbon, Portugal. pp.888-897, ⟨10.5220/0011895400003417⟩. ⟨hal-04045405⟩
39 Consultations
11 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More