Project Icon

PIA

Customizable Video Generation from Text-to-Image Models with Style Transfer

Product DescriptionPIA is a tool designed for personalized image animation using text-to-image models. It provides controlled motion, text-to-image alignment, and supports style transfer. Updates include Replicate API and Google Colab integration, and efficient image handling with scaled_dot_product_attention. It offers features like motion control, looping animations, and is accessible via platforms such as HuggingFace and OpenXLab.
Project Details