Alibaba's Qwen releases AI model that splits images into editable layers like PS (the-decoder.com)

Alibaba's AI unit Qwen has launched a groundbreaking image editing model, Qwen-Image-Layered, which allows users to split images into separate, editable layers, similar to Photoshop's functionality. This innovative model utilizes RGBA layers to enable independent editing of various image components, making tasks such as resizing, repositioning, and recoloring straightforward. Users can easily swap backgrounds, modify text, or manipulate objects without affecting the overall image integrity. The model offers flexibility by allowing images to be divided into either 3 or 8 layers, and users can further dissect each layer into additional components as needed. This release is significant for the AI/ML community as it bridges traditional image manipulation with structured, editable representations, opening up new possibilities for designers and content creators. Furthermore, Qwen has made the model's code available on GitHub and hosted it on platforms like Hugging Face and ModelScope, encouraging community engagement and collaboration. This initiative not only democratizes advanced image editing but also showcases the potential of AI in enhancing creative workflows, paving the way for future innovations in digital design.