The transformation of the video clip is stable diffusion. Experiment No. 1

The first experiment using Stable Diffusion is software based on neural networks. The reference from the previous post is taken as a basis. The essence of the processing: we cut the video into frames in AE, each frame or block of frames, depending on the scene, is passed through a neural network, where we get the generated image at the output in accordance with the specified description of the transformation. We assemble a sequence of frames into a video sequence, sew it together. For this video, it took more than 1000 images to be processed. The result is not perfect, but interesting)