Overcoming Data Limitation in Medical Visual Question Answering

Nguyen, Binh D, Do, Thanh-Toan ORCID: 0000-0002-6249-0848, Nguyen, Binh X, Do, Tuong ORCID: 0000-0002-6249-0848, Tjiputra, Erman and Tran, Quang D
(2019) Overcoming Data Limitation in Medical Visual Question Answering. In: MICCAI 2019, 2019-10-13 - 2019-10-17, Shenzhen, China.

[img] Text
1909.11867.pdf - Submitted Version

Download (441kB) | Preview


Traditional approaches for Visual Question Answering (VQA) require large amount of labeled data for training. Unfortunately, such large scale data is usually not available for medical domain. In this paper, we propose a novel medical VQA framework that overcomes the labeled data limitation. The proposed framework explores the use of the unsupervised Denoising Auto-Encoder (DAE) and the supervised Meta-Learning. The advantage of DAE is to leverage the large amount of unlabeled images while the advantage of Meta-Learning is to learn meta-weights that quickly adapt to VQA problem with limited labeled data. By leveraging the advantages of these techniques, it allows the proposed framework to be efficiently trained using a small labeled training set. The experimental results show that our proposed method significantly outperforms the state-of-the-art medical VQA. The source code is available at https://github.com/aioz-ai/MICCAI19-MedVQA.

Item Type: Conference or Workshop Item (Unspecified)
Depositing User: Symplectic Admin
Date Deposited: 18 Feb 2020 10:24
Last Modified: 24 Nov 2021 21:10
DOI: 10.1007/978-3-030-32251-9_57
Related URLs:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3075415