PixelProse
PixelProse is a large dataset for image-to-text tasks, ideal for training visual-language models.
What is PixelProse?
PixelProse is a comprehensive dataset with over 16 million image-text pairs generated using advanced models like Gemini 1.0 Pro Vision. It supports tasks such as image captioning and visual question answering, making it ideal for training and testing visual-language models.