Fetching the paper…

e-ViL: A Dataset and Benchmark for Natural Language Explanations in Vision-Language Tasks · Around