Adversarial VQA: A New Benchmark for Evaluating the Robustness of VQA Models
This addresses the robustness problem for VQA models in real-world scenarios, though it is incremental as it builds on existing VQA benchmarks.
The authors tackled the problem of evaluating the robustness of Visual Question Answering (VQA) models by introducing Adversarial VQA, a new benchmark collected via an adversarial human-and-model-in-the-loop procedure, and found that state-of-the-art models perform far worse on this benchmark compared to standard datasets, revealing their fragility.
Benefiting from large-scale pre-training, we have witnessed significant performance boost on the popular Visual Question Answering (VQA) task. Despite rapid progress, it remains unclear whether these state-of-the-art (SOTA) models are robust when encountering examples in the wild. To study this, we introduce Adversarial VQA, a new large-scale VQA benchmark, collected iteratively via an adversarial human-and-model-in-the-loop procedure. Through this new benchmark, we discover several interesting findings. (i) Surprisingly, we find that during dataset collection, non-expert annotators can easily attack SOTA VQA models successfully. (ii) Both large-scale pre-trained models and adversarial training methods achieve far worse performance on the new benchmark than over standard VQA v2 dataset, revealing the fragility of these models while demonstrating the effectiveness of our adversarial dataset. (iii) When used for data augmentation, our dataset can effectively boost model performance on other robust VQA benchmarks. We hope our Adversarial VQA dataset can shed new light on robustness study in the community and serve as a valuable benchmark for future work.