Towards robust visual question answering: Making the most of biased samples via contrastive learning
Models for Visual Question Answering (VQA) often rely on the spurious correlations, ie, the
language priors, that appear in the biased samples of training set, which make them brittle …
language priors, that appear in the biased samples of training set, which make them brittle …
Towards Robust Visual Question Answering: Making the Most of Biased Samples via Contrastive Learning
Q Si, Y Liu, F Meng, Z Lin, P Fu, Y Cao… - Findings of the …, 2022 - aclanthology.org
Abstract Models for Visual Question Answering (VQA) often rely on the spurious correlations,
ie, the language priors, that appear in the biased samples of training set, which make them …
ie, the language priors, that appear in the biased samples of training set, which make them …
Towards Robust Visual Question Answering: Making the Most of Biased Samples via Contrastive Learning
Q Si, Y Liu, F Meng, Z Lin, P Fu, Y Cao… - arXiv e …, 2022 - ui.adsabs.harvard.edu
Abstract Models for Visual Question Answering (VQA) often rely on the spurious correlations,
ie, the language priors, that appear in the biased samples of training set, which make them …
ie, the language priors, that appear in the biased samples of training set, which make them …