GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering
This provides a resource for advancing visual reasoning models, though it is incremental as it builds on existing VQA datasets with improved data generation and metrics.
The authors introduced GQA, a new dataset for visual reasoning and compositional question answering, addressing shortcomings of previous VQA datasets by using scene graphs to generate 22M diverse questions with functional programs for control and bias mitigation, resulting in human performance at 89.3% compared to 54.1% for strong VQA models.
We introduce GQA, a new dataset for real-world visual reasoning and compositional question answering, seeking to address key shortcomings of previous VQA datasets. We have developed a strong and robust question engine that leverages scene graph structures to create 22M diverse reasoning questions, all come with functional programs that represent their semantics. We use the programs to gain tight control over the answer distribution and present a new tunable smoothing technique to mitigate question biases. Accompanying the dataset is a suite of new metrics that evaluate essential qualities such as consistency, grounding and plausibility. An extensive analysis is performed for baselines as well as state-of-the-art models, providing fine-grained results for different question types and topologies. Whereas a blind LSTM obtains mere 42.1%, and strong VQA models achieve 54.1%, human performance tops at 89.3%, offering ample opportunity for new research to explore. We strongly hope GQA will provide an enabling resource for the next generation of models with enhanced robustness, improved consistency, and deeper semantic understanding for images and language.