Unable to run run_live_portrait.py on Google Colab
Opened this issue · 8 comments
In the Google Colab When I'm running this code
!python /content/Efficient-Live-Portrait/run_live_portrait.py -v /content/driving_video.mp4 -i /content/source.jpg -e -fp16
but it asking for run_live_portrait.py: error: the following arguments are required: -ref/--ref_image
--ref_image mean ?
Same as source image or first frame of the driving video where a clear face is available?
for the --ref_image image I tried with both same source image and first frame of the driving video where face is available
2024-07-22 14:53:18.480793999 [W:onnxruntime:, execution_frame.cc:879 VerifyOutputSizes] Expected shape from model of {12800,4} does not match actual shape of {8192,4} for output 451
.......................................................................................................................
.......................................................................................................................
2024-07-22 14:53:18.481856866 [W:onnxruntime:, execution_frame.cc:879 VerifyOutputSizes] Expected shape from model of {12800,1} does not match actual shape of {8192,1} for output 448
^C
But in both cases, the code stopped working.
My driving video was 1:1 format and in each frame the face is available
Ref image is a single face image you want to adapt for checking face id in the source image, source image should is the multiple faces image
I’ll update new code , for you can turn on/off the face id for easy running source
Hey @aihacker111 was the code updated? I got the same error: run_live_portrait.py: error: the following arguments are required: -ref/--ref_image
@PiovisTeam @NeuralFalconYT I'm still work faster for update new feature for animation with multiple face motion on multiple face on images, please don't test right now
@NeuralFalconYT Demo 2 motion in 1 image
img_1--test_driving_3_face.mp4
Thank you @aihacker111 :)
@NeuralFalconYT , @PiovisTeam New update is availabled, Please try it again and read more README