ESmike/true_imgs_modalities

Dataset Summary ESmike/true_imgs_modalities is a multimodal dataset containing 4,149 real images at a uniform resolution of 512 × 512, along with multiple derived modalities for each image. This dataset is intended for research and experimentation in computer vision, generative modeling, and multimodal learning. All images originate from real photographs curated and processed. Each base image is provided alongside additional modality representations (details in the Data Fields… See the full description on the dataset page: https://huggingface.co/datasets/ESmike/true_imgs_modalities.