I am looking for someone who could make me a bot that would generate and post tiktok videos, using artificial intelligence such as midjourney, chatgpt, elevenlabs, a speech to text ai and capcut for editing.
the bot should follow the instructions below::
TEXT:
-Transcribe audio from tiktok videos to text from a group of favorite videos on an account and make the text punctuate, I've seen AI's that could do this (a punctuated transcription of the audio) but I don't know which of them is the best so I thought maybe you would know better than me which one to choose
-ensure that I can add or remove favorite videos at any time
AUDIO :
-use ElevenLabs to create an audio from this text, I don't know yet which voice to use, I will tell you soon
IMAGES :
-Take the text and give it to chatgpt, adding a sentence (which I'll think about later) which will generate 20 prompts, with a line break for each new prompt.
-then go to midjourney on my discord account and enter these prompts one by one; to do this, go to a "newbies-*number*" channel and select "/imagine prompt" in the search bar, then enter the prompt
-download each image generated by these prompts in order,
-crop each image so that only the top-right image is visible (because Midjourney generates 4 images in 1 for each prompt), i.e. by cropping three-quarters of the main image, bearing in mind that each of the 4 images will be in 9:16 format
-download all the new cropped images
EDIT:
-create a new project on capcut
-select tiktok format (9:16)
-import the audio into this project
-then generate the automatic subtitles by going to "text" then "automatic captions"
-detect the word "num脙漏ro" (it芒聙聶s a french word, because the text will be in French) in the subtitles (this word will appear 4 times, ignore the third time), then if necessary separate the subtitle containing the word "num脙漏ro" so that the subtitle where this word appears begins with this word (if the subtitle containing this word already begins with this word, then there's no need to separate the subtitle)
-on the first appearance of this word, insert images no. 2,3,4,5,6,7,8, on the second appearance, insert images no. 2,9,10,11,12,13,14, on the fourth appearance, insert images no. 2,15,16,17,18,19,20, and cut all the previous images imported with the audio.
-before the first appearance of this word, insert image n脗掳1
-make image n脗掳2 last 1 second out of the 3 times it appears
-make sure image n脗掳1 appears until the first appearance of image n脗掳2
-make each of the 3 groups of 6 images between the different appearances of image n脗掳2 last until the appearance of the next image n脗掳2, and for the last group, make these last 6 images last until the end of the audio.
-within each of these 3 groups, ensure that all 6 images have the same appearance time
-assign an "animation" of the "combined" type to each of the 22 images (so that I can later define precisely which animation for each image) (to have an 芒聙聹animation芒聙聺, you just have to select the clip, then go to 芒聙聹animation芒聙聺, then 芒聙聹combined芒聙聺)
-so that all the following instructions apply to all subtitles at the same time and not just one, select a subtitle at random, then go to "style":
-then go to "style" (this is a different box to the previous "style" box) then to "text" then to "size" and select the value 13
-go to "font" then "English" and choose "Avenir next".
-go to "style" and select the third type of colorimetry (the yellow "T" surrounded by black) and change the color in the color palette so that the subtitle is yellow (just tell me how I can change the color myself to get precisely the one I want)
-go to "style" then "stroke", then select the black color, then "thickness" at the value 43
-go to "style" then 芒聙聹shadow芒聙聺, then select the color black, normally the values of the shadow characteristics are already good by default, but if this is not the case, select "opacity" at value 80, "blur" at value 35, "distance" at value 10, and "angle" at value -40
-also move the subtitles to the top third (show me how I can adjust this myself to place them where I want)
-also enlarge subtitles by selecting one at random and enlarging the block in which it appears on the video rendering by about 1.2 times its size (show me also how I can adjust this myself to enlarge them as I wish)
-upload video in 720p