Improving Data Augmentation for Robust Visual Question Answering with Effective Curriculum Learning
This work addresses inefficiencies in data augmentation for VQA models, offering an incremental improvement to existing methods.
The paper tackles the problem of redundancy in data-augmented training sets for visual question answering, which leads to inefficient training and compromised performance, by proposing an Effective Curriculum Learning strategy that trains models on easier samples first and removes less-valuable ones, resulting in enhanced performance with fewer training samples.
Being widely used in learning unbiased visual question answering (VQA) models, Data Augmentation (DA) helps mitigate language biases by generating extra training samples beyond the original samples. While today's DA methods can generate robust samples, the augmented training set, significantly larger than the original dataset, often exhibits redundancy in terms of difficulty or content repetition, leading to inefficient model training and even compromising the model performance. To this end, we design an Effective Curriculum Learning strategy ECL to enhance DA-based VQA methods. Intuitively, ECL trains VQA models on relatively ``easy'' samples first, and then gradually changes to ``harder'' samples, and less-valuable samples are dynamically removed. Compared to training on the entire augmented dataset, our ECL strategy can further enhance VQA models' performance with fewer training samples. Extensive ablations have demonstrated the effectiveness of ECL on various methods.