How to Make a Talking Animated AI Avatar With CapCut (With Mask)

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
in this video one of my subscribers comments on how to make AI talking animation without artlow since all YouTubers are just showing the same artlow version that looks terrible so I thought let's make a video about it also I'm quite bored with this one so why not make a new one that looks really really handsome and this one doesn't look handsome and this is something that you guys see instead of me so in this video I'm going to show you how to make this and this type of AI Avatar talking animation that I'm going to use personally in my channel not like other YouTubers who just tell you that they don't use that method to make their AI avatars it's all going to be free and easy to make so let's start I will show you guys two methods one with a mask and one without a mask first without a mask let me tell you one thing if you think that you know the first method and are going to skip it then don't do that because it intrigues the second method I'm just assuring you that if you watch the entire video you won't waste any time step one to generate an image there are penalty options out there pixel pixart canva C cut Lexica and many more I actually tried every single AI but no one gave me a satisfying result except Runway AI go to Runway AI then apply these settings and paste this prompt or your preferred prompt it depends on what you want to get also you can add some negative prompt if you want then click on generate image I tried every single option generated a bunch of images and then selected this image because I think it quite matches my criteria but it's not fully perfect so what we need to do now is click on this three dot then click on the image variation before that I tried image to image to explore this [Music] option but the results were so awful that is why don't go from image to image go for image variation now click on generate and keep trying until you find your character and remember that this is not our final image so if it's not so perfect then it's okay because the next Generation process is the main step where we will get the final image after generating three to four times my credit has been finished but I think this image came out really well though it is not our final image so don't worry about looks now download one of these images now go to Leonardo Ai and upload this image to here you might be thinking what if we directly generate our first image in Leonardo instead of Runway AI well I tried many times and the results are not acceptable as you can see after uploading this image click on image generation but the results are not that great right here I get an idea create a new account on Leonardo where you get Alchemy to use for free I already created a new account now apply these settings as I am showing [Music] here and paste The Prompt I will leave this prompt in the description so you can can try with it and click on generate and just from the same prompt you can see these results came out really well by just enabling Alchemy now if you want a more accurate result then you should try image to image right now what I'm going to do is upload the same image from Runway [Music] Ai and now click on [Music] generate and check out this result all of these results are looking great and you can literally go with anyone right but I want to go with this one because if you look carefully you can see this character is quite a little bit friendly whereas others are aggressive looking by their eyebrows and congratulations you have learned how to generate a perfect looking character for your faceless YouTube channel now here's the Second Step finding a background for your AI Avatar listen you can stick with the same background but you are not watching a regular tutorial if you watch me then you will create or make something above average the choice is yours by the way it's all about the result and that is God level go to Lexica AI I searched every single prompt so you don't have to just search for nighttime room View and pick one from here that is suitable for our [Music] studio I think this one is quite good or you can search exactly this prompt if you want to get this type of image also you can explore this type of image Style by clicking here after seeing many images I think it's better to stick with the first one after all First Love is something you want to forget but you can't so download any image that you like to use as your setup or Studio by the way we have completed this step step three now we need to remove the entire background except this table and our Avatar to do that go to remove.bg and click on change image background then just upload the image from which you want to remove the background in this case this is the image after completing the upload you can see the entire background has been removed perfectly um this little patch is not a big deal to remove now I don't know if I download it as it is it is going to have a transparent or white background so I put a color that is not in the image so I chose blue because I have seen it in Marvel film shooting then download it and go to canva for further editing select the YouTube thumbnail then upload the Avatar and background then drag it here and adjust the size next click on edit photo and select remove background you can skip the remove.bg site and directly remove it from canva I forgot this so whatever you do the results should be the same now click on Magic Eraser and you get a brush simply draw over the computer and it will automatically delete those parts now it's totally ready to place the background drag the background to the center move it to back and adjust it to how you want it to look just download this image decorating studio now if you are someone who doesn't like to decorate their setup or Studio then you can skip the next step go to Google and search for Adobe Firefly then go to generative fills or you can search for firefly.com upload in paint you will be directly redirected to this page now upload that image here adjust the position as everyone needs a laptop on their desk so draw that area where you want your laptop to be and write what you want for me it's a Macbook laptop click on generate as you can see I tried many times but these are not perfect because the desk is too small to place a laptop so we need to extend the desk to do that go to any uncrop website uncrop by clip drop is best anyway I go with neural love yeah you heard right love upload the canva edited image wait for a few seconds after complete processing and our image is ready now download it and go to canva to crop the image because in Firefly there is no option for cropping now upload that image to Firefly what we need to do is make the table complete and remove this paper so it looks [Music] clean now draw the area where you want your laptop to be describe what you want and click generate see the result and do what it needs to be perfect like increase the area edit the prompt or anything else click on more and keep [Music] generating I think this one I can keep you can decorate the rest of the blank space a wall clock air conditioner art piece lights Etc I need an art piece here so again follow the same [Music] step and it's done now download it go to canva and remove the [Music] logo editing image then upload this image to the photo P select magic lasso tool and start selecting the window [Music] area after selecting press delete key to delete the selection and click on the selected area to deselect it then one by one delete all the part of the window mistakenly I have a white background so it's looks white but it should be look like transparent like this and make sure you have no background in your image if you don't upload image in proper way then you could get in trouble in future editing process just like me but now I know that so I'm telling you and the proper way is don't click on new project then you will get this background instead click on upload from computer then select the image drag it or you can click on open next press Alt key and scroll up down to zoom in or zoom out the image then select healing tool and erase the mouth then save as PNG file finding background Now go to YouTube and search for nighttime ambient city views you can try any other search term that you want to put in your background I found this one and it's okay copy the link go to streamable because it's 42 minutes long paste the link trim the video and click on create video click on the three dots and click the download link and then download now we need a talking mouth animation download it from YouTube and never go into canva to remove the green screen it was terrible just go to cap cut PC for the rest of our editing create new project and upload all the stuff you [Music] need first drag this video to the timeline then the Avatar now adjust the video to show a specific part as a background then upload the mouth animation go to cut out check the green screen box pick this green color then and drag this slide bar to the highest crop the mouth animation then place this in the appropriate place and you're done don't rush to export take your time once you're satisfied export now check out the result I think it's pretty cool one thing to remember is to not add any effects or transitions unless you don't want to l lose video quality okay let's move on to our second method I swear you can't find this method on YouTube yeah you can see this method after I upload this video this method is a lot easier since we have already covered many steps step one image generation in first method we use Runway AI but you can skip that if you have Alchemy let me show you how first create a new account on Leonardo paste the same prompt and just edit and add a black mask select albo XL for Alchemy choose anime select 16.9 aspect ratio then generate as you can see these images came out really well but none of them are straight forward you can get that with many tries edit prompt and try but if you want an easy and fast method then try image to image click here turn on image input and upload the image I'm using the previous one because if I go to generate a new image then this video is going to be bigger so that would be a waste of time as you guys already know how to generate a good-looking AI Avatar then hit generate wait a few seconds and there you are and I'm absolutely surprised with the quality I don't think we can generate this type of highquality image with free tools I'm downloading this one editing image upload this image to photo and as you can see the background is black it is happened because we upload this image in image to image input so it is very easy to remove that just select the magic wand tool and click on the black spot then press delete for this area zoom in by pressing the alt plus scroll key in the mouse then press the H key to drag and adjust the position then select magnetic lasso and just draw the line it will automatically select the area then join the line and press delete now adjust the position to the face click magnetic lasso select the entire mask except the Border it depends on you I like the border to be there now select the brush tool and choose your preferred color from here and paint on the entire [Music] mask the black mask is looking good but it's your choice you're probably thinking why we need to do this well the answer you will get shortly I don't know why but I go for a grayish kind of color can you tell me in the comment which color is more preferable that I can use in my avatar no matter what color you choose you need to make it more realistic to do that select the smudge tool drag the end of the Mask area to the center from each side and drag one time from the top of the Mask now you can see the difference now click on the raster mask from the bottom right corner and then export the image first click on file then click on export as then choose PNG and save it then click on any blank area and press ctrl+ Z to undo and we got the whole image back again follow the same steps and Export the whole image editing video now open cap cut import those files I'm doing it on the same project I'm just replacing the Avatar image and deleting the mouth animation now Drag The Mask onto the timeline and set it in the image like this now use the mask to crop the mask image just carefully note what I am doing after setting up everything use the key frame and press the right arrow key 10 times then slide down the mask a little bit using Y AIS key to make it minus5 again press the right arrow key 10 times and reset the mask position to where it was just select this number and type zero again follow the same step move forward 10 times then select the Y AIS and type minus5 repeat it for the entire 10 seconds or 15 seconds or just cut it after repeating it for 5 to six times and make sure while you're cutting your video it should be cut from the place where a key frame is showing the mask down then press Arrow key 10 times and then cut your video otherwise it will not look perfect and then copy and paste it side by side now export [Music] this this is the ultimate outcome if you find it the most helpful and 100% original guide on how to make an AI Avatar then subscribe and you may actually send your AI Avatar to me on Instagram I want to see how it comes up the best top five will get a place on my community post see you in another video
Info
Channel: Smart Hustler
Views: 7,852
Rating: undefined out of 5
Keywords: ai talking avatar generator free, talking avatar for free, talking avatar, ai talking avatar, how to make talking ai avatar, how to create talking ai avatar, how to create a talking ai avatar for free, how to create ai avatar, how to create talking ai avatar free, how to make ai avatar for youtube videos, ai avatar, How to Make a Talking Animated Avatar with AI Tools, How to Make Talking AI Avatar Like @10xincome With CAPCUT(Secret Method), ai avatar generator free
Id: u0p8RZPWMpk
Channel Id: undefined
Length: 14min 56sec (896 seconds)
Published: Wed Mar 06 2024
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.