Cognitive resilience: Unraveling the proficiency of image-captioning models to interpret masked visual content

arXiv:2403.15876v1 Announce Type: new
Abstract: This study explores the ability of Image Captioning (IC) models to decode masked visual content sourced from diverse datasets. Our findings reveal the IC model's capability to generate captions from masked images, closely resembling the original content. Notably, even in the presence of masks, the model adeptly crafts descriptive textual information that goes beyond what is observable in the original image-generated captions. While the decoding performance of the IC model experiences a decline with an …

abstract arxiv capability captioning captions cognitive cs.ai cs.cv datasets decode diverse generate image image-captioning images resilience study type visual