VQA-CP v-2

'VQA-CP v2' is a dataset used to evaluate Visual Question Answering models by providing a set of questions and images designed to test their ability to generalize beyond biases present in training data.

Papers using VQA-CP v-2 (3)

Integrating Object Interaction Self-Attention and GAN-Based Debiasing for Visual Question Answering2025MPCAR: Multi-Perspective Contextual Augmentation for Enhanced Visual Reasoning in Large Vision-Language Models2025Overcoming Language Priors for Visual Question Answering Based on Knowledge Distillation2025