AiRethink

Realtime interactive programmable genAI video

With the Advanced Stream DIffusion based artist tool we built ( ivizDiff ) - we can use source images - and even use several and blend them. And prompts ( again many and wieght them) - and either live video or looped video files which we can control in realtime interactiveluy with all our advanced paramaters to make a realtime video with full interactive control including as you see here mapping live music, breath work, mic levels, hand gestures, and game controllers. We can also have several personally trained models LORAs on a users work or ideas. Here I am using my own original art work as image inputs and a lora trained on my style ( my art work) - all with me controlling my face video. See other demos of inputs from a living tree, or physically based manipulation for realtime architecture ideation.