I don't know the reason
https://youtu.be/Oi6OByxe5DM

After installation, it work fine, but has no Virtual Camera option on it, why?

I ' ve downloaded the desktop version,but when installing, it shows that can't extract file from
"installer://com.avatarify.desktop/0.10.0.0deploy_Qt_5_15_1_MSVC2019_64bit_full_torch.7z"
unable to open document"E:2//d3dcomplier_47>dll"

Please help me

The app worked great.

However, I am done using it and now want to uninstall. I tried uninstalling from the control panel, however that did not work. Please guide on how should I uninstall it. Just deleting the folder would do the job?

Thanks

When opening the EXE file, it simply does nothing.
I have a Ryzen 7 2700x and RTX 2070
I have open with admin as well as checked task manager and nothing shows forit, it just simply does not do anything, no error message, no nothing, kinda like a dud. Any ideas?

The app itself functions fine, with a visible working preview, but the output from Avatarify Camera is just a gray screen. Makes it pointless to use unless I use some sort of screen streaming software which was one of the biggest annoyances of the command line version.
I'm running it on a GTX 1070 Ti with an Intel i7 2600 and 20 GB of RAM, Windows 10 Build 19041.867

Hello, has anyone made it work with the RTX 30XX series?

Hello, I trained a first-order model dataset with a modified yaml config according to my purpose but I don't have any ideas on how to load these tar models here.

Hi! Very excited that you have made a windows compatible version of Avatarify! Thank you for that! I have the software installed, and have input faces I would like to use. The problem is that when I calibrate the selected face to my face no actual face movement of the mouth occurs. The eyes bug out and twitterpate, but no actual movement to sync up with mouth movement for speech. The whole face twitches and has mild distortion which looks very 'unrealistic.' This problem occurs also with all the "stock" faces that come with the software install.
I have a 1080TI cuda card.

stuck at solving environment

PlEASE CAN THIS WORK PERFECTLY ON NVIDIA GEFORCE 1060

I use my Play Station 3 camera through OBS virtual cam, but when I select OBS virtual camera in the options in the app it shows nothing in the preview, is OBS Virtual Cam not supported? My graphics card is RTX 2070S

I add my own image but found no way to remove them.

attached is how avatarify works on my PC i don't know if this is how it should work or maybe something is missing

I added a photo and I can't remove it from that faces bar. How do I remove it?

I know perfectly, that radeon gpu (in my case is rx 570) wont work in here and will be painfully slow, but maybe add some instructions on how to add cloud gpu rendering? I remember do it like that. i just dont want mess around with setting up avatarify itself, only with adding cloud render to your proggram

I have installed the Software normaly on my PC. I have the newest drivers for my rtx 3070. I have followed the instructions on the Page and my Webcam even turns on. I also get asked if wanted to grant CMD permission but still no Windoes opens. Any Ideas?
Thanks in advance.

Hi, I uninstalled Avatarify Desktop, but whenever I open video call programs they use Avatarify as the default camera. I tried looking for the driver in device manager but I cannot find it anywhere. How should I remove it from the list of the cameras? Please help. Thanks.

If i wear glasses it doesn't work

Stuck at solving environments on MSI stealth gaming pc, Nvidia rtx3060 8g ram windows 11. Any useful information will be appreciated

@JohanAR kindly drop the issue number or link to resolve this before subbing this issue as “duplicate”

Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: \

i need help about this please because the conda prompt is still loading over 10 minutes now

My computer configuration as below : Intel i5 [email protected], Nvidia 1080, 8GB memory,
everything works fine when I use a USB2.0 camera as video input, but when I switch to OBS camera, the avatars can not be driven by virtual video input, pls tell me what's going on ?

Hello! Running windows 10 with latest nvidia drivers on two machines.

One has two 3090s and the other has two a5000s.

I don't get anything to show up at all when i launch the app, but it displays as running in background on ask manager. Any idea what could be wrong?

Thanks!

what's this?any piecture to show?is that for android or ios platform?why the app is off in china?

I have an i7 pc with 32 gig ram and 1070 gtx
I installed it and the software starts correctly and I can choose the photos but it does not detect any camera...
I've tried Kinect & EOS & Fake Webcam but non working..
O by the way thanks for the answers

Hello， it working well one hour ago, but Avatarify Camera output nothing now.

Issue is back

I downloaded and ran the installer, The app appears on my desktop, but every time I try to run it I get this message. I have re-installed it multiple times.

bool AkVCamBridge::allocateDevice() {
m_ipcBridge.connectService(false);
m_ipcBridge.setDriverPaths({QDir::currentPath().toStdWString()});
auto devices = m_ipcBridge.listDevices();
std::cout << "List of devices:" << std::endl;
for (const auto &d : devices) {
std::cout << "\t" << d << std::endl;
}

bool isStarted{false};
if (devices.empty()) {
    qDebug() << "No available devices. Need to create one..." << supportedFormats.size();
    m_ipcBridge.deviceCreate(description, supportedFormats);
    devices = m_ipcBridge.listDevices();
    qDebug() << devices.size();
    if (!devices.empty()) {
        m_device = devices[0];
        qDebug() << m_device.c_str();
        for (int i = 1; i < 4; ++i) {
            QThread::sleep(4 * i);
            if ((isStarted = m_ipcBridge.deviceStart(m_device, format))) {
                break;
            }
        }
        if (!isStarted) {
            return false;
        }
    }
    //m_device = m_ipcBridge.deviceCreate(description, supportedFormats);
    //if (m_device.empty()) { // error during creating the device
    //    std::wstring_convert<std::codecvt_utf8<wchar_t>, wchar_t> converter;
    //    std::cout << "ipcBridge: " << converter.to_bytes(m_ipcBridge.errorMessage()) << std::endl;
    //    return false;
    //}
} else {
    qDebug() << "There are " << devices.size() << " devices. Choose 0...";
    m_device = devices[0];
    //m_ipcBridge.deviceEdit(m_device, L"Avatarify Camera", supportedFormats);
}
if (!isStarted) {
    isStarted = m_ipcBridge.deviceStart(m_device, format);
}
if (isStarted) {
    AmplitudeLogger::log("vcam_ok");
    std::cout << "Success initializing " << m_device << std::endl;
    return true;
} else {
    AmplitudeLogger::log("vcam_fail");
    std::cout << "Error initializing " << m_device << std::endl;
    std::wstring_convert<std::codecvt_utf8<wchar_t>, wchar_t> converter;
    std::cout << "ipcBridge: " << converter.to_bytes(m_ipcBridge.errorMessage()) << std::endl;
    return false;
}

}

Hi, amazing app works great, lot of fun to let Einstein talk haha! When is the output resolution going to be improved? Or are there already ways to improve quality or newer technologies to do similar things? Thanks Ali!

How to launch Avatarify.exe ?
It crashes with the APPCRASH error. (I'm running it on windows 7).
The file "Avatarify Desctop" in the folder c:\Users\Scar\Avatarify\ is missing.

Colab doesn't work either.

E5 2678v3 32GB RAM, RTX 2060
app runs, make calibration, when i click to image below - app crash witn no error.

I'm trying to transfer the pretrained model of first-order-model to libtorch model, but meet some problems.How did you obtain the .pt model?

Good afternoon Yesterday I installed Avatarify and when I turn it on... despite appearing in the task panel as running, the program screen does not appear.
I freaked out growing up that it was some computer problem, I uninstalled everything and did a clean install of windows. Install everything again, and Avatarify... and Not work...
I changed the graphics card for another old 580 and the application works (despite being so slow that nothing can be done...)I would like to know if anyone knows of a solution or if you plan to make an update to be able to use AVATARIFY with the 3060 cards.
It makes me very angry... It is an application that I really want to try.
Many thanks for everything!!

The app runs fine but is slow.
I checked in my windows Task manager, it's putting all the load on CPU and nothing on my nvidia 960M.
I'm on win 10.
I tried click right on the app name and running using graphics processor but that does not start the process.
If i double click on app, it start the process but it's too slow

Was anyone able to use it with omegle?

thanks for your wonderful job firstly. I wanna know if the project can be built and how to do that? I look forward to your reply as soon as possible.Thks

First of all, thank you for developing this project greatly.
avatarify-python works well as a container but being able to run it quickly on Ubuntu 21.04 would be a blast.
As you're also on Linux, I'm sure you understand.
Have a great day!

I was wondering what is the underlying model of this desktop version, is it the vox-adv-cpk.pth or another one? And is possible to replace the model easily with another one when running the app?

Is it possible to add support for 30XX graphics cards?

The tutorial says you can use an overlay slider to calibrate Avatarify, and in a screenshot there's an option to use Avatar as a virtual camera. I don't have the virtual camera option, nor do I have an overlay slider, and there isn't a virtual camera when I look at Discord.
Edit: Nevermind, it works now

AM FACING THE SAME ISSUE, WHEN I FIRST BOUGHT MY LAPTOP IT WAS WORKING PERFECTLY BUT ALL OF A SUDDEN IT STOP WORKING

The original python Avatarify had a zoom feature so you could get close to your face. Without this the software is crap. capture needs to be as close as possible to make it work good. Could someone add this feature? or is the code simple enough i could branch and do it myself? Would be great if the code just grabbed the face better and you didn't need zoom. either way it would improve this app a lot.

p.s. otherwise this is awesome release. Thanks!

When the application starts, it appears in the task manager, but does not start
VIDEO

I used google lab but still running very slow after using the picture its keeps lagging and face not in order, how can you please help me

Is there a version or scheme that runs on ios, or if I want to run on ios, what should I do?

how do we run this on google colab sir?