@Dark Web Red Room personally? No, I don't know many deep technical Mac users or administrators though there may be videos or tutorials going into this but I do not know even what to look for in the Mac world.
A relative of my wife’s has end stage ALS where the muscles cannot move at all so the only way she can communicate is through an eye movement based computer which is made for ALS patients. Before the patient gets to end stage and can still speak they sample their voice so that later when they can’t speak the computer system speaks for them in their own voice and I think this tech is what they use. ALS is a terrible disease and is terminal with no cure. Having the ability to communicate, especially with small children who need to hear their mom’s voice, is really a big plus of this type of technology.
Please follow these steps: 1. record a good sample 2. tell the boss you're working from home from now on 3. have the computer respond with "have you tried turning it off and on again?" to incoming phone calls from the office 4. Randomly interchange with "sorry we don't support Macs" or "you're going to need to purchase a new computer" 5. Sit back and drink beers while watching TV. 6. Profit!
We know absolutely _nothing_ about coding and python, but we _really_ would like to try this out. Is there any way someone can give a step-by-step tutorial that's a little easier for us to understand?
@@WIDMTV yeah and we mostly see his face but whats important is not shown with the excuse" but i already done it so i wont show u how to do it because yes " ... so suck it " .....
@@xXWhatzUpH8erzXx I'm in a IT carrier now, and what I was saying is, that , this tool might be so complex to execute for ppl that knows nothing about scripting/programming or things like dat
It would make more sense for them to make a voice changer, where it takes a recording of one person saying something, and makes it sound like someone else. Rather than using text to speech. Then you could "act" a little in the recording, to try to keep the emphasis and inflection and sound less like a robot.
Here's a program I found that is supposed to do what you stated in your comment and they have a demo here on UA-cam and the company is voiceconverter.net but I've only messed with it a little and have not had much success converting voices and there's not a lot of support but I guess I have to mess with it more. Here's the UA-cam demo link : ua-cam.com/video/4Uzf0SL7rjA/v-deo.html
@@BrutusJones The link doesn't work anymore. I really want to try it out. Do you know where I can get it or can you kindly send the install pack directly to me.
@@nalineepruttanopajai3437 Since this video, there's been a lot more video's on the subject. Search in UA-cam for voice cloning but here's one link that seems pretty good but there might be better ones. You just have to do some digging. ua-cam.com/video/I-tpsOSy5l4/v-deo.html
there was an isaac asimov tv show on abc in the late 80s called “probe” where the main character who was supposed to be a genius built a work-in-progress program from scratch that was able to clone a voice and foil the bad guys over a p.a. system. roughly 30 years later and the capability is finally getting to the masses. awesome stuff but we’re finally getting there.
Well, the "Yes" is only for using Audicy to cut a message together. This doesn't need a "Yes" to make "you" say something, just a long enough recording of you saying whatever you do. Then "you" might be saying: "Yes, I do agree with you about closing down my account and sending all money to the account of Anders Jackson.". :-)
sadly some of the required packages arent even available anymore. The biggest issue i have with ML is all the cool demos dont work because they are outdated and unmainted
Thanks For The Cloned Card My Dogg *DAWGGOAT* I Almost Went Crazy Cuz Of Yo Card😂 When Atm Approved The Card And Start Throwing Da Amount I Click On Out.. You Just Got Yourself Another Long Term Business Dude Brrr
You made a high quality video, but it was though for me to follow your instructions when I couldn't see what you were doing on the computer. Also, you have MAC instead of Windows. It's not your fault, it was another issue for me to understand though. I insist: awesome video, even though it wasn't that clear for me.
pip install - r requirements.txt worked for me (remove space between - and r ) -- pip install r requirements as he said it did not. Also "switching over" to the download directory is done by the command cd
At first I couldn't get it to work on my Windows 10, but now it works. Every software package involved needs to have the right version. *CUDA Toolkit 10.0 (Sept 2018)* developer.nvidia.com/cuda-10.0-download-archive *cuDNN v7.6.5 (November 5th, 2019), for CUDA 10.0* (you must create an account there to get it) developer.nvidia.com/cudnn *PyTorch, old enough for CUDA 10.0* pip install torch==1.2.0 torchvision==0.4.0 -f download.pytorch.org/whl/torch_stable.html
That’s amazing. If it’s like that now, in a short while it will be almost perfect. Once it’s working properly, you can make a movie with characters from the past😱
And nowadays with the leaps in text to image generation it won‘t be many years until generated novels, movies and shows, with characters of your choice (even yourself and your friends) are good enough to be worthwhile.
What do you mean? Are you making a better/easier program than this one? Please release it with easy installation/easy way to train it to become better overall/for a specific if so when done.
You don't need an NVIDIA GPU to get it working. I've got it running on an Intel CPU, and it's slow, but it works! I can't make a tutorial on how to install it on Intel CPU's/GPU's, because it's *_WAY_* too complicated to install. Literally everything requires tons of effort to install.
@@pabitrashahi45 I spent about a month trying to get it working, there's no way I'm doing it again just for a tutorial. Plus, my computer isn't powerful enough to be recording and running that program at the same time, so the footage would be broken (and possible corrupted) and unusable. And finally, I'd have troubles even editing the video, again, because my computer is not very powerful.
Hi. How are you? Does CorentinJ (the real time voice cloning) work on a mac running on windows bootcamp and with a intel graphics 6000? I would be super grateful with any info you have. Please. Hope you have a good day
@novaspiritTech Hello I would like to know if I can use this software for professional business purpose. I will develop it according to the steps shown, my only concern is that are there any copyrights to this software. I would really appreciate a reply.
cant we just go back to the good old days where everything was in an easy to use .exe file? because if pip install is the future, then i dont want the future
im doing a cg short film, i really want this to make the voices, trying to clone seiju voice? or something that works at least. The moments when character scream or yell or cry i just can edit audio from animes.
Is there a project that supports cloning voice timbre without TTS? I've tried Real-Time-Voice-Cloning but it seems to generate only TTS text to the target voice. I would like to record my own phrase with my voice to keep emotions and inflections, and then encode it as if spoken by the target person's voice.
dude i never even touched python and the stuff you need, i work as a game dev, ( designer) confident to sday that most people give up after 1h of not even getting it to work.
Did you meet this error Traceback (most recent call last): File "demo_cli.py", line 3, in from synthesizer.inference import Synthesizer File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/inference.py", line 1, in from synthesizer.tacotron2 import Tacotron2 File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/tacotron2.py", line 1, in from synthesizer.utils.text import text_to_sequence File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/utils/text.py", line 2, in from . import cleaners File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/utils/cleaners.py", line 14, in from unidecode import unidecode ModuleNotFoundError: No module named 'unidecode' after that I installed unidecode by pip install unidecode but I still get the same error
@@sarabadawi For me, after running sudo python3 demo_cli.py, the last line in the terminal is a prompt that asks, "Reference voice: enter an audio filepath of a voice to be cloned (mp3, wav, m4a, flac, ...): " Do you see that? If so, you can upload a .wav file (.mp3 failed for me for some reason) and give it the path after than prompt. It will then process that file then prompt you for the text you want to generate speech for. After entering that, you'll get an output .wav file that you can download. Don't be disappointed though...I still can't get anything that really sounds as good as the examples yet.
you know it makes me curious what troubles you could run into when using voice cloning for film projects. Like say I decided to use a sample of a celebrity's voice, would they be able to sue me for using their voice? Or could it somehow be deemed fair use somehow? [Granted I wouldn't actually risk using a celeb's voice in my own projects since I'd rather not risk getting sued up the ass for using their voice]
showed me this when trying to install requirements.txt " ERROR: Could not find a version that satisfies the requirement tensorflow-gpu=1.10.0 (from -r requirements.txt (line 1)) (from versions: none) ERROR: No matching distribution found for tensorflow-gpu=1.10.0 (from -r requirements.txt (line 1)) " is there a way to fix this issue? I'm on a mac.
if you are using anaconda, use the code "conda install -c anaconda git" you can download and install anaconda if you don't have it installed in you system
I would use it to incorporate as the method to create responses in an automated responses in a PBX like FreePBX. Replace Lenny whose scripts have been around for a while.
I'm trying to follow along because this is the only video I could find that "teaches" you how to get this running, but he says things like, "Now open up a terminal" or "change over to the folder where you downloaded it to" but I don't know what he means by change over, was there supposed to be a folder I could open? and where was he typing??????? I really just want to get this working, is there a more beginner-friendly video somewhere?
hes on linux, I tried to get it to work on windows and its quite fucky, even after installing all the dependencies and stuff, even cuda 10 ans cudnn, I didnt get it work cause for some reason it couldn't find it . It's really weird on Windows, if someone can make a version taht goes from installation to getting to work that would be really nice
Hey Don! This would be perfect for what I want to do but couldn't get it work on my computer. Got everything downloaded but when I put commands into the command prompt window all I get is "...not recognized as an internal or external command..." and from what I've researched fixing this is beyond my computer skills. Any chance if setting this up as one of your repositories on gethub so it is not so complicated. I've looked on line to see if there is software for sale that does the same thing but couldn't find anything. I am not interested in cloing my voice. I want to clone other voices using wave files that I already have. Let me know if there is anything for sale that could do this.Thanks,Jim.Nordand, Wa
My terminal stops the process because this file is missing. %userprofile%\AppData\Local\Program Files\Python\Python37\lib\site-packages\torch\_C could put the whole process on google drive to facilitate installation on windows 10?
I may sound dumb but I keep getting "ERROR: No matching distribution found for tensorflow-gpu=1.10.0 (from -r C:\python37\Real-Time-Voice-Cloning-master2 equirements.txt (line 1))" even after I manually installed it. I would love to see the WHOLE process. pip install -r requirements.txt does not work
Hey I fixed it, what helped was when I installed a free trial of pycharm, which, whenever I ran the project in it, it would tell me another plugin I needed to install or it install it for me, and it works fine now
I spent all day trying to get it to work on Windows with Anaconda and with PIP but its just over my head. Running nvidia, have cuda and cudnn, have pytorch installed, but I get errors ImportError: DLL load failed: The specified module could not be found. or something about numpy
olá amigo, no windows estou com problema (ModuleNotFoundError: No module named numba.core), quando vou iniciar python demo_toolbox.py. teria alguma solução? \ eu consigo mandar os arquivos de audio pelo prompt, porem não consigo abrir toolbox.py
@@NovaspiritTech doesn't sound anything like it. I'd like to figure ot how they're really doing this because I've heard incredible examples, but this implementation sounds like robotic crap, not even close to the papers examples. Any idea why?
@bmt79 you have to use old tensorflow and make sure you have all the requiments loaded some of the requirements are old use the following. tensorflow==1.15 umap-learn visdom librosa>=0.5.1 matplotlib==2.0.2 numpy>=1.14.0 scipy>=1.0.0 tqdm sounddevice SoundFile Unidecode inflect PyQt5 multiprocess numba==0.48
I am having issues installing the requirements. For some reason installing tensorflow 1.15 it says for me: "could not find a version that satisfies the requirement tensorflow==1.15". Is there any way to solve this? I am also running pip3 on 20.0.2
This is very cool, I have couple of questions: 1. Is there a pre-built version for Windows 10? (like they did with the DeepFake different apps versions). 2. Will this work with other languages other than English? if so is there a specific list of languages supported? Thanks ahead and keep up the good work! :)
How does CorentinJ's project compare with his resemble.ai.. Lyrebird, or Baidu's deepvoice3 / pytorch ? And how difficult would it be to add real-time voice recognition/dictation for your own voice, and then map it to output a cloned voice all in real-time? Do you recommend any open-source SpeechRecognition project? Finally, any thoughts on Anand Pawara's Real_Time_Image_Animation that uses pytorch and a "First Order Motion Model" to animate an image in real-time (similar to 4dface / eos)?!
2:02 - "open up a terminal". What is a terminal? Windows commandl line does not understand these commands. Terminal of Python 3.8 (32bit) - the same. What a terminal we are talking about?
i'm running this using anaconda. After all the installations i tried the test configuration code "python demo_cli.py" and it showed me this C:\Users\user\Documents\Real time voice cloning\Real-Time-Voice-Cloning\encoder\audio.py:13: UserWarning: Unable to import 'webrtcvad'. This package enables noise removal and is recommended. warn("Unable to import 'webrtcvad'. This package enables noise removal and is recommended.") Arguments: enc_model_fpath: encoder\saved_models\pretrained.pt syn_model_dir: synthesizer\saved_models\logs-pretrained voc_model_fpath: vocoder\saved_models\pretrained\pretrained.pt low_mem: False no_sound: False seed: None no_mp3_support: False C:\Users\user\anaconda3\envs\REAL\lib\site-packages\librosa\core\audio.py:162: UserWarning: PySoundFile failed. Trying audioread instead. warnings.warn("PySoundFile failed. Trying audioread instead.") Librosa will be unable to open mp3 files if additional software is not installed. Please install ffmpeg or add the '--no_mp3_support' option to proceed without support for mp3 files. Please how do I install the webrtcvad or bypass this stage
it is easy to install and works fine on my computer but it doesn't uses the latest versions of libaries like tensoreflow so u have to the perticular one
I got it running but honestly I think it's pretty bad, I found one on another one that is ran off the cloud and I tried to run it off my computer but couldn't get it to work...anyway that one specifically creates nearly perfect voice clones.
This is not real time. Real time would be if you talk into a mic and it immediately transforms your voice in *real time*, instead of clicking and waiting for it to process.
I just spent over seven hours trying to install the software on Windows and Linux but there's just way too many bugs in the install scripts/broken packages, incompatibilities etc. I'm super frustradted, and I've given up by now. DO NOT try it, it's just not worth your time!
im wondering if anyone can give me a hand here, im trying to use the software but i always get large gaps and silences in the audio output and it will read the words i type but it never sounds anything like the voice samples
For Windows, you should open the PowerShell instead of Terminal. You can open it by right-click on the Windows icon at the bottom-left corner and select Windows PowerShell. Is there anyone who can give me the following file? My terminal stops the process because this file is missing. %userprofile%\AppData\Local\Programs\Python\Python37\lib\site-packages\torch\_C
i cloned joshua's voice from war games!! LOL
twitter.com/NovaspiritTech/status/1173738560832331776
Novaspirit Tech, Pay for my VPN.
I want that voice for espeak, any idea how I could do that? Given it's a synthesis to start with.... That's so cool!
@Dark Web Red Room Don may be able to give you some pointers, I've not done anything with it myself yet unfortunately.
@Dark Web Red Room I don't know specifically but these is a shell of sorts on a Mac, I am not a Mac guy myself so I couldn't advise you unfortunately
@Dark Web Red Room personally? No, I don't know many deep technical Mac users or administrators though there may be videos or tutorials going into this but I do not know even what to look for in the Mac world.
A relative of my wife’s has end stage ALS where the muscles cannot move at all so the only way she can communicate is through an eye movement based computer which is made for ALS patients. Before the patient gets to end stage and can still speak they sample their voice so that later when they can’t speak the computer system speaks for them in their own voice and I think this tech is what they use. ALS is a terrible disease and is terminal with no cure. Having the ability to communicate, especially with small children who need to hear their mom’s voice, is really a big plus of this type of technology.
It's easy to forget how useful this technology can actually be Thank you for providing a good example
Please follow these steps:
1. record a good sample
2. tell the boss you're working from home from now on
3. have the computer respond with "have you tried turning it off and on again?" to incoming phone calls from the office
4. Randomly interchange with "sorry we don't support Macs" or "you're going to need to purchase a new computer"
5. Sit back and drink beers while watching TV.
6. Profit!
IT crowds ??
Google has A.I that can do that lmao
Roy would like to hire you...
I’m the 100th liker of this comment XD 😂
Invest
If anyone can make a complete install video for windows, I would definetly watch, but theres none : /
I would not refuse such a video!) All my attempts have failed.
@Sba Stuff But i have one... gtx960 to be clear. And still nothing)
@Sba Stuff ? :0
@@yori4434 poorlydocumented.com/2019/11/installing-corentinjs-real-time-voice-cloning-project-on-windows-10-from-scratch/
@@yori4434 Really? I got it working fine
every time I put a file it says
exception:
help
the same for me. anyone figure this out?
And this is how you get that anime girl voice to your virtual assistant
We know absolutely _nothing_ about coding and python, but we _really_ would like to try this out. Is there any way someone can give a step-by-step tutorial that's a little easier for us to understand?
Nah script community don't give a shit about ppl that can actually give this a real use...they just don't share
same! I downloaden everything... but no idea, opened up python, typed what the guy said.... nothing !
@@WIDMTV yeah and we mostly see his face but whats important is not shown with the excuse" but i already done it so i wont show u how to do it because yes " ... so suck it " .....
@@Nekotico All of the code in this repo is open sourced, it's all being shared for free.
@@xXWhatzUpH8erzXx I'm in a IT carrier now, and what I was saying is, that , this tool might be so complex to execute for ppl that knows nothing about scripting/programming or things like dat
It would make more sense for them to make a voice changer, where it takes a recording of one person saying something, and makes it sound like someone else. Rather than using text to speech. Then you could "act" a little in the recording, to try to keep the emphasis and inflection and sound less like a robot.
Here's a program I found that is supposed to do what you stated in your comment and they have a demo here on UA-cam and the company is voiceconverter.net but I've only messed with it a little and have not had much success converting voices and there's not a lot of support but I guess I have to mess with it more. Here's the UA-cam demo link : ua-cam.com/video/4Uzf0SL7rjA/v-deo.html
@@BrutusJones the website dont work
@@BrutusJones The link doesn't work anymore. I really want to try it out. Do you know where I can get it or can you kindly send the install pack directly to me.
@@nalineepruttanopajai3437 Since this video, there's been a lot more video's on the subject. Search in UA-cam for voice cloning but here's one link that seems pretty good but there might be better ones. You just have to do some digging. ua-cam.com/video/I-tpsOSy5l4/v-deo.html
Wow, I think we're going to get some brand new Peppa Pig episodes.
New Mlp Ytp episodes.
you are a fu*king genius
@@mlgproplayer2915 more YTP
there was an isaac asimov tv show on abc in the late 80s called “probe” where the main character who was supposed to be a genius built a work-in-progress program from scratch that was able to clone a voice and foil the bad guys over a p.a. system. roughly 30 years later and the capability is finally getting to the masses. awesome stuff but we’re finally getting there.
Build pre-trained models using your own voice? Man, it really checks your patience had to give up after 15 days...
I hope someone continued the development of this. I haven't got it working, it's very difficult, but I'm hoping it gets some changes.
Curious if this is what scammers are using to clone peoples voices?
I heard you shouldn't say yes when they ask you questions.
Well, the "Yes" is only for using Audicy to cut a message together.
This doesn't need a "Yes" to make "you" say something, just a long enough recording of you saying whatever you do.
Then "you" might be saying: "Yes, I do agree with you about closing down my account and sending all money to the account of Anders Jackson.". :-)
sadly some of the required packages arent even available anymore. The biggest issue i have with ML is all the cool demos dont work because they are outdated and unmainted
Thanks For The Cloned Card My Dogg *DAWGGOAT* I Almost Went Crazy Cuz Of Yo Card😂 When Atm Approved The Card And Start Throwing Da Amount I Click On Out.. You Just Got Yourself Another Long Term Business Dude Brrr
Soooooo... grab a snippet of Arnold Schwarzenegger man!!! Holy cow.
You made a high quality video, but it was though for me to follow your instructions when I couldn't see what you were doing on the computer. Also, you have MAC instead of Windows. It's not your fault, it was another issue for me to understand though. I insist: awesome video, even though it wasn't that clear for me.
pip install - r requirements.txt worked for me (remove space between - and r ) -- pip install r requirements as he said it did not. Also "switching over" to the download directory is done by the command cd
plz can you explain more how to setup correctly for windows 10 ??? thanks
@fortnite4201 x WHy the hell does a voice synthesizer, which is audio, need a certain high performance graphics card. Makes no sense. So annoying!
@@phily8093 because there's an artificial intelligence software running with CUDA as it's primary processing
At first I couldn't get it to work on my Windows 10, but now it works. Every software package involved needs to have the right version.
*CUDA Toolkit 10.0 (Sept 2018)*
developer.nvidia.com/cuda-10.0-download-archive
*cuDNN v7.6.5 (November 5th, 2019), for CUDA 10.0* (you must create an account there to get it)
developer.nvidia.com/cudnn
*PyTorch, old enough for CUDA 10.0*
pip install torch==1.2.0 torchvision==0.4.0 -f download.pytorch.org/whl/torch_stable.html
Oh and I have to run *cmd* as administrator.
If you use this together with video deep fakes, you'll create a very authentic clone video of whomever you choose.
so, you can clone voice in other language?
i am' Braziliam, i Want to clone some voices an make them speak portuguese
LOL
That’s amazing.
If it’s like that now, in a short while it will be almost perfect.
Once it’s working properly, you can make a movie with characters from the past😱
And nowadays with the leaps in text to image generation it won‘t be many years until generated novels, movies and shows, with characters of your choice (even yourself and your friends) are good enough to be worthwhile.
As a side note, you can also try out Resemble AI for cloning your own voice free 😁
I'm having trouble with that I've been trying to figure it out
I am a comp sci student currently working on the same problem with a different solution, I'll post my results soon hopefully
do a windows 10 instalation video. this seems impossible to even get to work :(
What do you mean? Are you making a better/easier program than this one? Please release it with easy installation/easy way to train it to become better overall/for a specific if so when done.
You don't need an NVIDIA GPU to get it working. I've got it running on an Intel CPU, and it's slow, but it works! I can't make a tutorial on how to install it on Intel CPU's/GPU's, because it's *_WAY_* too complicated to install. Literally everything requires tons of effort to install.
why you can't make tutriol video on it you got good not of views if you make it and upload it on your channel because it's demand is high
@@pabitrashahi45 I spent about a month trying to get it working, there's no way I'm doing it again just for a tutorial. Plus, my computer isn't powerful enough to be recording and running that program at the same time, so the footage would be broken (and possible corrupted) and unusable. And finally, I'd have troubles even editing the video, again, because my computer is not very powerful.
@@pabitrashahi45 I'm not going to make a tutorial on that. It is way too complicated, and it doesn't work very well.
Hi. How are you?
Does CorentinJ (the real time voice cloning) work on a mac running on windows bootcamp and with a intel graphics 6000?
I would be super grateful with any info you have.
Please.
Hope you have a good day
I feel like this would really work well with someone like the fact core, due to speaking in a completely monotone voice.
@novaspiritTech Hello I would like to know if I can use this software for professional business purpose. I will develop it according to the steps shown, my only concern is that are there any copyrights to this software. I would really appreciate a reply.
This software generates a voice characteristic profile, and a good use for it would be voice authentication
cant we just go back to the good old days where everything was in an easy to use .exe file? because if pip install is the future, then i dont want the future
If anyone is having trouble and stuck at all the warnings in the demo scripts, run it as administrator
If you try it and get some Module Errors try to execute in the current path *pip install -r requirements.txt* . This will install the modules
im doing a cg short film, i really want this to make the voices, trying to clone seiju voice? or something that works at least. The moments when character scream or yell or cry i just can edit audio from animes.
how did you exported the voice from toolbox? recorded screen and converted to mp3?(like I do) or wrote a code?
I cant get this to work. I copy and paste the Pytorch code in the command prompt but it gives me syntax error or gives me error messages help!!!
Is there a project that supports cloning voice timbre without TTS?
I've tried Real-Time-Voice-Cloning but it seems to generate only TTS text to the target voice. I would like to record my own phrase with my voice to keep emotions and inflections, and then encode it as if spoken by the target person's voice.
I'm looking for the same thing. Did you find anything ?
dude i never even touched python and the stuff you need, i work as a game dev, ( designer) confident to sday that most people give up after 1h of not even getting it to work.
Did you meet this error Traceback (most recent call last):
File "demo_cli.py", line 3, in
from synthesizer.inference import Synthesizer
File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/inference.py", line 1, in
from synthesizer.tacotron2 import Tacotron2
File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/tacotron2.py", line 1, in
from synthesizer.utils.text import text_to_sequence
File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/utils/text.py", line 2, in
from . import cleaners
File "/home/ubuntu/voice/Real-Time-Voice-Cloning/synthesizer/utils/cleaners.py", line 14, in
from unidecode import unidecode
ModuleNotFoundError: No module named 'unidecode'
after that I installed unidecode by pip install unidecode
but I still get the same error
I have this error too, you are not alone! I hope we hear a fix
run this and try again "pip3 install unidecode" or all the package "pip3 install -r requirements.txt"
I ran sudo python3 demo_cli.py and it works now how can I test this without GUI I only use the terminal can anyone know?
@@sarabadawi For me, after running sudo python3 demo_cli.py, the last line in the terminal is a prompt that asks, "Reference voice: enter an audio filepath of a voice to be cloned (mp3, wav, m4a, flac, ...): "
Do you see that? If so, you can upload a .wav file (.mp3 failed for me for some reason) and give it the path after than prompt. It will then process that file then prompt you for the text you want to generate speech for. After entering that, you'll get an output .wav file that you can download.
Don't be disappointed though...I still can't get anything that really sounds as good as the examples yet.
you know it makes me curious what troubles you could run into when using voice cloning for film projects. Like say I decided to use a sample of a celebrity's voice, would they be able to sue me for using their voice? Or could it somehow be deemed fair use somehow? [Granted I wouldn't actually risk using a celeb's voice in my own projects since I'd rather not risk getting sued up the ass for using their voice]
Voice isn't really copyrighted, like you could do a impression.
What if you have zeveral older GPU's? I have 3 GTX640' that re passivly cooled. Can I gang up the Kuda Cores?
showed me this when trying to install requirements.txt
" ERROR: Could not find a version that satisfies the requirement tensorflow-gpu=1.10.0 (from -r requirements.txt (line 1)) (from versions: none)
ERROR: No matching distribution found for tensorflow-gpu=1.10.0 (from -r requirements.txt (line 1)) "
is there a way to fix this issue? I'm on a mac.
I've got it working after 4 hours but now says AssertionError: Torch not compiled with CUDA enabled? any ideas on a fix?
0:30 Was "Reasonable" from a different recording after "The pricing is very"? It sounds like 2 different cuts that were stuck together?
Perhaps it wasn't reasonable the first time - And then the economy got better and he didn't have the time
What app are you using to open the terminal ? am lost.
Im on windows. I open python and paste the Git clone command but nothing happens
same... it only worked when I switched to mac...
Make sure you have git installed and you arent in the python shell, use the windows cmd, powershell or the git shell
if you are using anaconda, use the code "conda install -c anaconda git"
you can download and install anaconda if you don't have it installed in you system
Hi! Great video!! Do you know if this software also works in french?
I would use it to incorporate as the method to create responses in an automated responses in a PBX like FreePBX. Replace Lenny whose scripts have been around for a while.
Can i feed it with more data and leave training for longer to get better results?
I'm trying to follow along because this is the only video I could find that "teaches" you how to get this running, but he says things like, "Now open up a terminal" or "change over to the folder where you downloaded it to" but I don't know what he means by change over, was there supposed to be a folder I could open? and where was he typing??????? I really just want to get this working, is there a more beginner-friendly video somewhere?
hes on linux, I tried to get it to work on windows and its quite fucky, even after installing all the dependencies and stuff, even cuda 10 ans cudnn, I didnt get it work cause for some reason it couldn't find it .
It's really weird on Windows, if someone can make a version taht goes from installation to getting to work that would be really nice
@@aethernets9442 Well shit.
@@YeraTheNewbie Yeah Ive been trying to do the same. If anyone knows a windows 10 tutorial thatd be much appreciated.
you can put yourself in the lower-right coner, not full screen... :D
This kind of software is going somewhere
Hey Don! This would be perfect for what I want to do but couldn't get it work on my computer. Got everything downloaded but when I put commands into the command prompt window all I get is "...not recognized as an internal or external command..." and from what I've researched fixing this is beyond my computer skills. Any chance if setting this up as one of your repositories on gethub so it is not so complicated. I've looked on line to see if there is software for sale that does the same thing but couldn't find anything. I am not interested in cloing my voice. I want to clone other voices using wave files that I already have. Let me know if there is anything for sale that could do this.Thanks,Jim.Nordand, Wa
I think you might not have pip installed
My terminal stops the process because this file is missing. %userprofile%\AppData\Local\Program Files\Python\Python37\lib\site-packages\torch\_C could put the whole process on google drive to facilitate installation on windows 10?
I may sound dumb but I keep getting "ERROR: No matching distribution found for tensorflow-gpu=1.10.0 (from -r C:\python37\Real-Time-Voice-Cloning-master2
equirements.txt (line 1))" even after I manually installed it. I would love to see the WHOLE process.
pip install -r requirements.txt does not work
hey im having the same problem, did you ever fix it? thanks
same, fix?
Hey I fixed it, what helped was when I installed a free trial of pycharm, which, whenever I ran the project in it, it would tell me another plugin I needed to install or it install it for me, and it works fine now
@@caltic4743 thanks I'll try that!
Looks cool, gonna have to try it out. Thanks Don
I spent all day trying to get it to work on Windows with Anaconda and with PIP but its just over my head. Running nvidia, have cuda and cudnn, have pytorch installed, but I get errors
ImportError: DLL load failed: The specified module could not be found.
or something about numpy
why not using a proper microphone to have some quality results?
His microphone sounds good to me
olá amigo, no windows estou com problema (ModuleNotFoundError: No module named numba.core), quando vou iniciar python demo_toolbox.py. teria alguma solução? \ eu consigo mandar os arquivos de audio pelo prompt, porem não consigo abrir toolbox.py
Those who are having trouble installing it can look for the Google Colab version
so this works only for mac right?
Absolutely nothing I do will get this program to work. *sigh*
Totally off topic but what is the name of the song in your intro?
I have an 8gb AMD RX480 card, not a GForce with Cudo... Can I use this software? I mean my computer is no slouch, 32GB ram, Ryzen 5 3rd Gen.
@3:53 what do u mean by appropiate location?
where should we exactly extract pretrained models?
can u pls tell
He already showed you. You just put it in the clonder directory.
does this only work for english sentences or also other languages?
i'm not sure, i think for other languages you will need to train the model
@@NovaspiritTech please let us know how to train model for other languages
When I try to run the demo_cli.py it i get: ,,ModuleNotFoundError: No module named 'unidecode'" How can I fix it?
@Sijin John I got: ModuleNotFoundError: No module named 'encoder.inference' on MAC OS
Do these softwares also allow you to extract the voice from an existing video, learn it, then use that voice to replace one in another video?
The intro music is nice :) Well produced.
Just curious,when you say open up a terminal, which program is that?
10:45 Nothing will change, because you could have used the same approach before for banks with simple voice recording.
One of the hardest voices to replicate has been Joshua from War Games. It would be interesting to see how that sounds.
Imma try it now
i did it!!! twitter.com/NovaspiritTech/status/1173738560832331776
@@NovaspiritTech doesn't sound anything like it. I'd like to figure ot how they're really doing this because I've heard incredible examples, but this implementation sounds like robotic crap, not even close to the papers examples. Any idea why?
Amazing how close it is to original voice!!
What version of python were you using at the time of this video?
I really want to use this to make memes
I put python demo_toolbox.py
and nothing happens
@bmt79 you have to use old tensorflow and make sure you have all the requiments loaded some of the requirements are old use the following.
tensorflow==1.15
umap-learn
visdom
librosa>=0.5.1
matplotlib==2.0.2
numpy>=1.14.0
scipy>=1.0.0
tqdm
sounddevice
SoundFile
Unidecode
inflect
PyQt5
multiprocess
numba==0.48
Now the question is how you turned your now digitally-cloned-voice into a TTS program?
I am having issues installing the requirements. For some reason installing tensorflow 1.15 it says for me: "could not find a version that satisfies the requirement tensorflow==1.15".
Is there any way to solve this? I am also running pip3 on 20.0.2
This is very cool, I have couple of questions:
1. Is there a pre-built version for Windows 10? (like they did with the DeepFake different apps versions).
2. Will this work with other languages other than English? if so is there a specific list of languages supported?
Thanks ahead and keep up the good work! :)
Quite rough sounding but interesting none the less
How does CorentinJ's project compare with his resemble.ai.. Lyrebird, or Baidu's deepvoice3 / pytorch ?
And how difficult would it be to add real-time voice recognition/dictation for your own voice, and then map it to output a cloned voice all in real-time? Do you recommend any open-source SpeechRecognition project?
Finally, any thoughts on Anand Pawara's Real_Time_Image_Animation that uses pytorch and a "First Order Motion Model" to animate an image in real-time (similar to 4dface / eos)?!
There is a guy with a deep country voice that acts and does voice advertising. I would love to use his voice. Lol
When you type something, can you show your screen instead of yourself? That would help a lot. Thanks for the video :)
2:02 - "open up a terminal". What is a terminal? Windows commandl line does not understand these commands. Terminal of Python 3.8 (32bit) - the same. What a terminal we are talking about?
i'm running this using anaconda.
After all the installations i tried the test configuration code "python demo_cli.py" and it showed me this
C:\Users\user\Documents\Real time voice cloning\Real-Time-Voice-Cloning\encoder\audio.py:13: UserWarning: Unable to import 'webrtcvad'. This package enables noise removal and is recommended.
warn("Unable to import 'webrtcvad'. This package enables noise removal and is recommended.")
Arguments:
enc_model_fpath: encoder\saved_models\pretrained.pt
syn_model_dir: synthesizer\saved_models\logs-pretrained
voc_model_fpath: vocoder\saved_models\pretrained\pretrained.pt
low_mem: False
no_sound: False
seed: None
no_mp3_support: False
C:\Users\user\anaconda3\envs\REAL\lib\site-packages\librosa\core\audio.py:162: UserWarning: PySoundFile failed. Trying audioread instead.
warnings.warn("PySoundFile failed. Trying audioread instead.")
Librosa will be unable to open mp3 files if additional software is not installed.
Please install ffmpeg or add the '--no_mp3_support' option to proceed without support for mp3 files.
Please how do I install the webrtcvad or bypass this stage
it is easy to install and works fine on my computer but it doesn't uses the latest versions of libaries like tensoreflow so u have to the perticular one
I got it running but honestly I think it's pretty bad, I found one on another one that is ran off the cloud and I tried to run it off my computer but couldn't get it to work...anyway that one specifically creates nearly perfect voice clones.
can you tell me the name of the one in the cloud please?
@@wiltango8509 check my channel, I have an entire walk-through on it.
This is not real time. Real time would be if you talk into a mic and it immediately transforms your voice in *real time*, instead of clicking and waiting for it to process.
What window manager are you using?
Looks like i3 tiling window manager
Yo, someone should try to see if you can replace aku's new voice actor with aku's old actor using this type of software and see how it works
Ugh.. looks like I need a new computer in order to do cool shit.
I am getting error with the installing requirements step. Something about "could not find a version that satisfies the requirement tensorflow-gpu".
same here, did you ever fix it?
@@caltic4743 I did. Something to do with tensorflow version I believe. Just have to use a different version.
Yic17 Studio thanks man I will try to do that
@@Yic17Studio looks like i fixed it with "pip install PyQt5", just needed that to work thanks again
I just spent over seven hours trying to install the software on Windows and Linux but there's just way too many bugs in the install scripts/broken packages, incompatibilities etc. I'm super frustradted, and I've given up by now. DO NOT try it, it's just not worth your time!
Is there any probability to have background noise at duplicate voice clip ?
im wondering if anyone can give me a hand here, im trying to use the software but i always get large gaps and silences in the audio output and it will read the words i type but it never sounds anything like the voice samples
ua-cam.com/video/b1fzyM0VhhI/v-deo.html this is easier and so much better
Nice sir,,,,, how can I insert/match a my own new word in download stylish audio voice/ascent?
I thinking of using this for temporary voice acting for my animation project
Does this only support English ?
AMAZING! Can I clone a voice that speaks Italian language?
For Windows, you should open the PowerShell instead of Terminal. You can open it by right-click on the Windows icon at the bottom-left corner and select Windows PowerShell.
Is there anyone who can give me the following file? My terminal stops the process because this file is missing.
%userprofile%\AppData\Local\Programs\Python\Python37\lib\site-packages\torch\_C
What Terminal do you use at around 2 minutes in?