Stable Diffusion Animation Create Youtube Shorts Dance AI Video (Tutorial Guide)

hey everyone lately I've been creating YouTube shorts Dance videos using our Channel's AI Covergirl Nancy we're making use of stable diffusions and root face swap techniques in today's tutorial I'm going to walk you through the process and steps of how I did it some of you have been asking how I create these Dance videos so here we go believe it or not some folks questioned if it's an AI and someone even asked for this girl's Tik Tok ID well the twist is she's entirely virtual some of you in the comments claim that the move to move extension doesn't work with version 1.6 I'm here to show you that it indeed works I know some of you are asking in a good manner but those jerks are trying to ruin the comment sections well yes this video are going to slap those jerks face let's kick things off with automatic 1111 the easiest way to create animations using stable diffusions is is by using move to move it doesn't require extensive preparation like other methods you only need your short videos drag and drop them add your prompts and generate animations I'm also using the face swap extension based on the rot model it's important to note that this extension is uncensored and NSFW enabled I know some of you like that huh links to everything I mention will be in the description below the SD web reactor extension has been a smooth experience for me during installation and initial use you don't need to download additional model files as with the RP web UI the first time you run the extension it downloads the necessary model files the extension also supports Google collab SD web UI before you use the phes swap extension make sure to disable the SD web UI RP extension and restart automatic 1111 both extensions cannot run simultaneously it'll crash the system now let's dive into the move to move tab in this updated version they've added a feature called Movie Editor you can select a specific frame and add it to the move to move settings write a text prompt for that frame it allows you to change a scene from the original video like turning a person sitting in a park into someone sitting in a futuristic spaceship this feature makes a huge difference compared to using deorum or stable diffusion e synth utility uh uh to create unique AI animations for example you uh can drag and drop a video and use the movie editor to select a frame number you can mark down the frame numbers and create specific text prompts for each the clip interrogate key frame and deep buru key frame features allow you to play around with the key frame image this extension is more convenient than absin utility which requires extracting all image frames from a video clip and doing image to image for each key frame with move to move you just type a text prompt on the selected keyframe number in my example I've selected a key frame and added my text prompt about a woman dancing the clip interrogate key frame and deep baru key frame download the custom model the first time you use them while waiting for the downloads we're going to enable the reactor extension below much like the RP web UI extension select the face image you want for the face swap and click enable additional settings are available if you have multiple people in the video clip and the face swap image allowing you to specify the target people this feature can handle about three people I'll keep the other settings as default let's try a new video for face swap and move to move animation this time since the white dress dancing videos were already generated with nany's face we'll test it with different denoising strength and noise multipliers for this demonstration uh the first generation is a bit rough with a cartoony style using rev animated checkpoint with denoising strength set to 0 5 we're using realistic Vision checkpoint models aiming for something similar to the YouTube short videos but with the realistic Vision checkpoint models and lower denoising strength it significantly improves stable diffusion is finishing the generation and I'll be back with the results all right the command prompt has stopped and it's 100% done let's see the results on the right side you see the face swap result with row point4 the noising strength and noise multiplier set to one I've also tested other settings to reduce noise and flickering which I'll show at the end of the video so there you have it it's a tutorial on using the stable diffusion extension purely for generating animations without using other AI tools let's check it out so this is the 0.4 result with noise multiplier one uh as you can see there's still a little flickering on the hair and the background and here's the better one she as you can see is way smoother not much flickering it just like that YouTube short video that I posted that's all for today's tutorial I hope you've gained some inspiration on how to create realistic animation video clips similar to the ones I've posted on YouTube shorts if you haven't already subscribed to our Channel please click the button now until next time see you in the next video have a fantastic day

Channel: Future Thinker @Benji

Views: 250,037

Keywords: stable diffusion, stable diffusion tutorial, ai animation, AI-generated animations, YouTube short videos, denoising settings, realistic animations, stable diffusion img2img, ai art, install stable diffusion tutorial, move-to-move extension, AI animations, Stable Diffusions, virtual characters, YouTube shorts, dance videos, Roop Face Swap, mov2mov, animation stable diffusion, animation stable diffusion tutorial

Id: lrFcvq-BI84

Length: 8min 2sec (482 seconds)

Published: Thu Oct 12 2023