How to Make Talking AI Avatar (Lip Sync AI): From Beginner to Expert

in this video I am going to show you all the steps to create a talking AI Avatar with lips sync using an ultra secret tool the best part is that it's really simple and completely free I will also uncover some very useful tips and tricks to make this process Flawless and super easy so the first step is to generate an avatar for your video for this I will be showing you Leonardo AI it has a lot of features and it's a great tool now the best way to generate images would be to ask chat GPT to generate a prompt for you however personally I just scroll through and copy The Prompt of the images that I like and modify it a bit I like the image of this woman and if you want this prompt you can get it from the description below now this result is from my first attempt and I'm pretty happy with it so we will use this as an example now to change the place where she is we will use fukus open it and then click on the advanced check boox in the settings tab tab change the aspect ratio to 16x 9 so it's suitable for YouTube and increase the number of images to 10 so you have more options to choose from there's also a tab for style we can mix and match these different styles to get far different results I am going to use only fucus V2 and SII digital art then in the model tab make sure that Jugger not XL base model is selected I will use this prompt which you can find in the description below then we check on input image go to the image prompt Tab and check advanced so it displays additional options click on face swap and drop your image there increase the stop at to 0.95 to make it keep the pose and increase the weight to 0.8 or 0.9 to increase the similarity of the images click on generate and wait for the images let's review the 10 images as you can see the hands are one of the most difficult things to do for this model so try to keep them out or verify that they look good I will use this AI avatar for the animation before proceeding please leave a like And subscribe it helps me a lot and allows me to create more content like this now we need a script for this Avatar choose topics you would like the Avatar to talk about and use chat GPT or perplexity AI to generate a script for you chat GPT is the default that most people use but perplexity AI is also good and doesn't require a login it searches the web in real time and provides up-to-date information on various topics I recommend trying both and selecting the answer you prefer then we head over to get the voiceover we will use 11 Labs it is a great AI voiceover platform and has a good forever- free plan the paid plan though is a good option and is on promotion at the time of filming this I use this tool all the time for all my videos the link is in the description below after signing in you will see this page in the voices tab explore section you can find multiple voices to use try them and find the one that you like most then in the speech tab create section you will find the speech synthesis where you can Define if it's text to speech and the settings I will use Joanne she has a soft and lovely voice let's preview her voice everyone thinks of changing the world but no one thinks of changing himself now change the voice settings to your liking I will use the default settings the next step is to select the model I will use 11 multilingual V2 to my liking it's the one that delivers the best results and it allows you to write in 29 different languages write the script that you want her to say and click on generate it will take a few seconds let's listen to her hello my name is Joanne and I am an AI Avatar I enjoy discussing AI tools and how to make money using them if you are interested in these topics please subscribe to my channel great she is lovely finally we are ready to animate the AI Avatar you generated we will use sad talker to animate this Avatar image with just a few clicks sad talker is super simple and easy to use the great thing about this project is there are multiple options where you can run this there is a free Google collab a free hugging face spaces you can also add this as an extension in your stable diffusion web UI I will walk you through all of these options as well as show you how you can further improve the results generated by sad talker first let's look at the free hugging face spaces this is a free demo running on hugging face you can simply click here to upload your image and upload your own audio here I recommend you change the face model resolution to 512 for better resolution set the pre-process to full so it does not crop our image check the still mode so it does not move a lot and use gfp Gan as face enhancer that will improve the quality end result of your face the problem here is that sometimes the application is too busy and it won't allow you to generate you might need to try it different times to make it work next I want to show you how to run this inside Google collab simply click on this go to Google collab and make sure that you go and create a copy in your own drive next go to runtime and make sure that GPU is selected as the hardware accelerator now let's quickly go over the code of what is happening here first you are simply looking at what type of GPU is being used the second block of code installs all the required packages including python as well as the packages required to run sad talker after that the next block of code is actually downloading the required model files the next part of the code takes care of the images that are being uploaded if you want to upload your image you will need to upload the image into this folder by drag and drop we also need an audio file for that we're going to to be using this other folder driven audio drag and drop your audio there make sure that the image is PNG or it won't work once we run the next block of code you will see this drop- down menu where you can select what image you want to use the last step is the animation here is where you must change the file name to the one that you uploaded now we can simply run this block of code this is going to take a while depending on how long your audio is this step is optional preview and download your anim you can run this step to easily visualize and download your video but if you don't want to run this step you can find your video in the results folder lastly I will show you how to run this on your stable diffusion web UI this is the easiest way to do it and it's the way I use all the time to install sad Talker go to the extensions tab then click on install from URL paste this URL it will be in the description click on install then go to the installed tab make sure that it's enabled and click click apply and restart UI after reloading go to the sad talker Tab and upload your image and audio into the corresponding areas we are not changing the pose because that's a good one but you can try different pose Styles until you find the one that you think is the best for the face model resolution we will go with 512 to improve quality for the pre-process we will use full I think it's the best one and it does not crop your image normally when you use full you should enable still mode the batch size generation is to improve the speed that your GPU will render the video normally 2 is good I have tried higher and it crashes lastly use the gfp Gan enhancer to improve video quality and click on generate when it finishes you will be able to preview it and download it hello my name is Joanne and I am an AI Avatar I enjoy discussing AI tools and how to make money using them if you are interested in these topics please subscribe to my channel in the next video I am going to show you how to do it with another tool and we are going to compare the results if you're interested in more content like this please leave a like And subscribe to my channel I appreciate your support a lot thank you for watching

Channel: MintedWealth

Views: 24,729

Keywords: how to make talking ai avatar, lip sync ai, how to create talking ai avatar, talking avatar, how to create ai avatar, ai talking avatar, talking avatar for free, how to make ai avatar, talking ai avatar, lip sync avatar, talking avatar ai, free talking avatar ai, lip sync ai avatar, ai avatar with lip sync, ai lip sync, ai lip sync free, lip sync ai tutorial, free lip sync ai, talking animated avatar, how to make talking avatar, sadtalker talking avatar ai, Ai avatar

Id: 0772Fce9YLQ

Length: 8min 17sec (497 seconds)

Published: Sat Feb 03 2024