We observe a severe under-reporting of the different kinds of errors that Natural Language
Generation systems make. This is a problem, because mistakes are an important indicator of …

Referenceless metrics (eg, CLIPScore) use pretrained vision--language models to assess
image descriptions directly without costly ground-truth reference texts. Such methods can …

The automatic evaluation of image descriptions is an intricate task, and it is highly important
in the development and fine-grained analysis of captioning systems. Existing metrics to …

Recent advances in image captioning are driven by increasingly larger-scale vision--
language pretraining, relying on massive computational resources and increasingly large …

Alan Turing's pioneering vision of machines in the 1950s, that are capable of thinking like
humans is still what Artificial Intelligence (AI) and Deep Learning research aspires to …

Earlier research has shown that evaluation metrics based on textual similarity (eg, BLEU,
CIDEr, Meteor) do not correlate well with human evaluation scores for automatically …

Automatic evaluation metrics hold a fundamental importance in the development and fine-
grained analysis of captioning systems. While current evaluation metrics tend to achieve an …

Image captioning models are typically trained by treating all samples equally, neglecting to
account for mismatched or otherwise difficult data points. In contrast, recent work has shown …

Tasks that require modeling of both language and visual information, such as image
captioning, have become very popular in recent years. Most state-of-the-art approaches …

Finding true words that describe an image seems easy, but it is finding the right words for an
image that poses a communicative challenge. Describing an image to make it accessible for …