Open ssinha89 opened 3 years ago
Sorry I have not test the code. The face_tracker is another file. I will update this later today. I am preparing the 3D animation part. That part is large so I am figuring out how to upload to github.
Thanks for releasing the data preprocessing code. However, I am unable to run the single_video_preprocess.py file due to the following error: ModuleNotFoundError: No module named 'face_tracker'
Also it would be helpful if the details on generating the 3d animation video aligned_ani.mp4 was provided, as it is mentioned that the animation video is needed when the --warp_ani flag is set.
Please visit https://github.com/1adrianb/face-alignment to install face_alignment package.
@ssinha89
In sigle_video_preprocess.py
command = 'ffmpeg -framerate 25 -i ./temp%05d'%config.batch_id + '/%05d.png -vcodec libx264 -vf format=yuv420p -y ' + cropped_video_path
^
SyntaxError: invalid syntax
how to deal with this error?
single_video_preprocess.py
TAB string exists, replace it with four spaces. It's ok.
modify 70
line, as follows:
try:
try:
_crop_video(original_video_path, config.batch_id, 1)
except:
print('some error when crop images.')
except:
pass
Sorry I have not test the code. The face_tracker is another file. I will update this later today. I am preparing the 3D animation part. That part is large so I am figuring out how to upload to github.
@lelechen63 Thanks, looking forward to the details on preparing the input 3d animation video. I am wondering how to evaluate the entire pipeline model on audio input? Currently only the model for evaluating head pose from audio input is provided, could you also share demo model and code for deriving facial expressions on landmarks from input audio? The demo code for texture generation takes landmarks as input, however it is not clear from the existing code how to generate the landmarks from input audio.
The face_tracker is another file. I will update this later today.
Hey, any updates on this? Really looking forward to trying it
The face_tracker is another file. I will update this later today.
Hey, any updates on this? Really looking forward to trying it
Check the readme in data folder.
Check the readme in data folder.
I did, I don't see the face_tracker file or any workarounds. I'm running the following command.
python single_video_preprocess.py --extract_landmark --video_path=../demo/00181_aligned.mp4
And here's the list of packages I've installed. Am I doing something wrong?
$pip list
Package Version
--------------- ---------
certifi 2020.6.20
cycler 0.10.0
decorator 4.4.2
dlib 19.21.0
dominate 2.5.2
face-alignment 1.1.0
imageio 2.9.0
kiwisolver 1.2.0
matplotlib 3.3.1
networkx 2.5
numpy 1.19.1
opencv-python 4.4.0.42
Pillow 7.2.0
pip 19.0.3
pyparsing 2.4.7
python-dateutil 2.8.1
pytorch-msssim 0.2.1
PyWavelets 1.1.1
scikit-image 0.17.2
scipy 1.5.2
setuptools 40.8.0
six 1.15.0
tifffile 2020.8.25
torch 1.4.0
torchvision 0.5.0
tqdm 4.48.2
Updated.
From: Dipam Vasani notifications@github.com Sent: Saturday, August 29, 2020 3:34 PM To: lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion Talking-head-Generation-with-Rhythmic-Head-Motion@noreply.github.com Cc: Chen, Lele l.chen@rochester.edu; Mention mention@noreply.github.com Subject: [EXT] Re: [lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion] Error in single_video_preprocess.py (#4)
Check the readme in data folder.
I did, I don't see the face_tracker file or any workarounds. I'm running the following command.
python single_video_preprocess.py --extract_landmark --video_path=../demo/00181_aligned.mp4
And here's the list of packages I've installed. Am I doing something wrong?
$pip list Package Version
certifi 2020.6.20 cycler 0.10.0 decorator 4.4.2 dlib 19.21.0 dominate 2.5.2 face-alignment 1.1.0 imageio 2.9.0 kiwisolver 1.2.0 matplotlib 3.3.1 networkx 2.5 numpy 1.19.1 opencv-python 4.4.0.42 Pillow 7.2.0 pip 19.0.3 pyparsing 2.4.7 python-dateutil 2.8.1 pytorch-msssim 0.2.1 PyWavelets 1.1.1 scikit-image 0.17.2 scipy 1.5.2 setuptools 40.8.0 six 1.15.0 tifffile 2020.8.25 torch 1.4.0 torchvision 0.5.0 tqdm 4.48.2
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_lelechen63_Talking-2Dhead-2DGeneration-2Dwith-2DRhythmic-2DHead-2DMotion_issues_4-23issuecomment-2D683333279&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=hXfJF45WmBPeioP1nVK6ZCHVdV7IhM3-HcEF7YV7wdQ&s=wlSEMw0P6VwJhOR7KXuC2jni9DWP7bd_Ztd_HCUvi7o&e=, or unsubscribehttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AF4TJXDEGXMII45W7UEBHPDSDFJ3HANCNFSM4QK6SUDQ&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=hXfJF45WmBPeioP1nVK6ZCHVdV7IhM3-HcEF7YV7wdQ&s=E1cp4-4d7BwPi9tQySNPnXBzatA77TgxBJqTSM-rEAA&e=.
Thanks, getting another error now
ModuleNotFoundError: No module named 'utils'
and if I do
pip install utils
it gives me
ModuleNotFoundError: No module named 'utils.util'
Can you do a quick pip freeze > requirements.txt
and upload the requirements.txt
? That would be really helpful.
Hi there,
The code has not been tested yet. I will test it later. I will aslo upload utils file later.
Thanks.
Best, Lele
From: Dipam Vasani notifications@github.com Sent: Saturday, August 29, 2020 3:44 PM To: lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion Talking-head-Generation-with-Rhythmic-Head-Motion@noreply.github.com Cc: Chen, Lele l.chen@rochester.edu; Mention mention@noreply.github.com Subject: [EXT] Re: [lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion] Error in single_video_preprocess.py (#4)
Thanks, getting another error now
ModuleNotFoundError: No module named 'utils'
and if I do
pip install utils
it gives me
ModuleNotFoundError: No module named 'utils.util'
Can you do a quick pip freeze > requirements.txt and upload the requirements.txt? That would be really helpful.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_lelechen63_Talking-2Dhead-2DGeneration-2Dwith-2DRhythmic-2DHead-2DMotion_issues_4-23issuecomment-2D683334137&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=WlZoInBWE_dSVPdF3zCveGpS72fvrsw357MUD-v_VHE&s=3h3kW9ZSyYFP06FvJK-C3MEhQBokpRoLUR2abpGuGFE&e=, or unsubscribehttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AF4TJXE7WBE7PMNXLQ3CKE3SDFLAHANCNFSM4QK6SUDQ&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=WlZoInBWE_dSVPdF3zCveGpS72fvrsw357MUD-v_VHE&s=cRqGltBaz6ouJod_quEwWddBr7dcAjEQHFPDh_W_Y9M&e=.
@lelechen63 would be helpful to have a google colab for the demo, that way there are less environment issues
@lelechen63 getting the same error as @dipam7. Would appreciate utils file. Thanks
@lelechen63 Having same issue as @dipam7 as well. Utils file would be great, cheers!
you can rename 'util' folder to 'utils'
Thanks, getting another error now
ModuleNotFoundError: No module named 'utils'
and if I do
pip install utils
it gives me
ModuleNotFoundError: No module named 'utils.util'
Can you do a quick
pip freeze > requirements.txt
and upload therequirements.txt
? That would be really helpful.
you can rename 'util' folder to 'utils'
It should be good to go. I already tested all the functions in single_video_preprocess.py.
@lelechen63 Thanks for updating the 3d animation data preparation code. It would also be great if you could share the details of the expression generation from audio as mentioned here.
@lelechen63 Thanks, looking forward to the details on preparing the input 3d animation video. I am wondering how to evaluate the entire pipeline model on audio input? Currently only the model for evaluating head pose from audio input is provided, could you also share demo model and code for deriving facial expressions on landmarks from input audio? The demo code for texture generation takes landmarks as input, however it is not clear from the existing code how to generate the landmarks from input audio.
@ssinha89 were you able to get this working, do you have a google colab notebook for the demo?
@lelechen63 FileNotFoundError: [Errno 2] No such file or directory: '../basics/standard.npy'
in single_video_preprocess.py
@lelechen63
FileNotFoundError: [Errno 2] No such file or directory: '../basics/standard.npy'
insingle_video_preprocess.py
uploaded
@lelechen63 ValueError: The passed save_path is not a valid checkpoint: ./Data/net-data/256_256_resfcn256_weight
, I have downloaded the PRN trained model at GoogleDrive
, and put it into Data/net-data.
I will fix it tomorrow.
Sent via the Samsung Galaxy S10, an AT&T 5G Evolution capable smartphone Get Outlook for Androidhttps://aka.ms/ghei36
From: Pearson Zhao notifications@github.com Sent: Monday, August 31, 2020 12:31:11 AM To: lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion Talking-head-Generation-with-Rhythmic-Head-Motion@noreply.github.com Cc: Chen, Lele l.chen@rochester.edu; Mention mention@noreply.github.com Subject: [EXT] Re: [lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion] Error in single_video_preprocess.py (#4)
@lelechen63https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_lelechen63&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=l2xUJO22u6Yj8SItiqv8KAqO9xBTGnb7hTuQssNq0Sw&s=A966iVW3n11djyVAj8fsVKLcKCRdiQu79tE0POwD2zw&e= ValueError: The passed save_path is not a valid checkpoint: ./Data/net-data/256_256_resfcn256_weight, I have downloaded the PRN trained model at GoogleDrive, and put it into Data/net-data.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_lelechen63_Talking-2Dhead-2DGeneration-2Dwith-2DRhythmic-2DHead-2DMotion_issues_4-23issuecomment-2D683542845&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=l2xUJO22u6Yj8SItiqv8KAqO9xBTGnb7hTuQssNq0Sw&s=BP8pKth5Ju3EBHgfiQMfzxY2R6sstnJxod6gnM8NN1s&e=, or unsubscribehttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AF4TJXECZCAODHRYA4IAICTSDMRQ7ANCNFSM4QK6SUDQ&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=qRrg4VTSFIYKRMz88LmmxufThi8eOutCmC64X_B4jVc&m=l2xUJO22u6Yj8SItiqv8KAqO9xBTGnb7hTuQssNq0Sw&s=7dcJdOaCQS8Vdr-ZcwKiRkm0r7Axr8OmmftowUhVUXE&e=.
I have solved from https://github.com/clks-wzz/PRNet-Depth-Generation/issues/3
, In Data/net-data
add 256_256_resfcn256_weight.index file. I don’t know if this is the case.
@lelechen63
File "find_camera.py", line 2, in
@lelechen63 if direction.ndimension() == 1: AttributeError: 'bool' object has no attribute 'ndimension'
. I have downloaded SoftRas
project, and python setup.py install
in conda envs
.
python 3.6.10
pytorch 1.1.0
cuda 10.0.130
cudnn 7.6.5
I tested my video, and what caused this error? Thanks.
@lelechen63 TypeError: __init__() got an unexpected keyword argument 'camera_up'
, camera_up
parametrer unnecessary?
@lelechen63
TypeError: __init__() got an unexpected keyword argument 'camera_up'
,camera_up
parametrer unnecessary?
I got the same error. On removing that argument, I got the error
File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/transform.py", line 65, in forward vertices = srf.look(vertices, self._eye, self.camera_direction) File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/functional/look.py", line 31, in look if direction.ndimension() == 1: AttributeError: 'bool' object has no attribute 'ndimension'
@lelechen63
TypeError: __init__() got an unexpected keyword argument 'camera_up'
,camera_up
parametrer unnecessary?I got the same error. On removing that argument, I got the error
File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/transform.py", line 65, in forward vertices = srf.look(vertices, self._eye, self.camera_direction) File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/functional/look.py", line 31, in look if direction.ndimension() == 1: AttributeError: 'bool' object has no attribute 'ndimension'
You need to use my softrs to install it. Instead of the official one. You need to cd to my softrs folder and run the installation.
@lelechen63 Tanks,are all the codes updated? I can't import soft_renderer
@lelechen63 Tanks,are all the codes updated? I can't import soft_renderer
Yes. All the functions in single_video_process.py have been tested. You need to go to softrs folder to install it.
@lelechen63
TypeError: __init__() got an unexpected keyword argument 'camera_up'
,camera_up
parametrer unnecessary?I got the same error. On removing that argument, I got the error File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/transform.py", line 65, in forward vertices = srf.look(vertices, self._eye, self.camera_direction) File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/functional/look.py", line 31, in look if direction.ndimension() == 1: AttributeError: 'bool' object has no attribute 'ndimension'
You need to use my softrs to install it. Instead of the official one. You need to cd to my softrs folder and run the installation.
The SoftRas folder inside data/face-tools is empty. How to run your version of softras?
@lelechen63
TypeError: __init__() got an unexpected keyword argument 'camera_up'
,camera_up
parametrer unnecessary?I got the same error. On removing that argument, I got the error File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/transform.py", line 65, in forward vertices = srf.look(vertices, self._eye, self.camera_direction) File "/usr/local/lib/python3.7/dist-packages/soft_renderer-1.0.0-py3.7-linux-x86_64.egg/soft_renderer/functional/look.py", line 31, in look if direction.ndimension() == 1: AttributeError: 'bool' object has no attribute 'ndimension'
You need to use my softrs to install it. Instead of the official one. You need to cd to my softrs folder and run the installation.
The SoftRas folder inside data/face-tools is empty. How to run your version of softras?
Please git clone the softras folder from https://github.com/goddice/SoftRas
@lelechen63 I've got an error when try to import your version: ModuleNotFoundError: No module named 'soft_renderer.cuda.load_textures'
Thanks for releasing the data preprocessing code. However, I am unable to run the single_video_preprocess.py file due to the following error: ModuleNotFoundError: No module named 'face_tracker'
Also it would be helpful if the details on generating the 3d animation video aligned_ani.mp4 was provided, as it is mentioned that the animation video is needed when the --warp_ani flag is set.