The ControlNet Pose tool is crafted to generate images featuring the same pose as the individual in the input image. It utilizes Stable Diffusion and Controlnet methods to duplicate the weights of the neural network blocks into both a “locked” and “trainable” copy. Users can customize various parameters such as the number of samples, image resolution, guidance scale, seed, eta, added prompt, negative prompt, and resolution for detection. The prediction process generally completes in approximately 21 seconds.
Try now