【自己幫影片換主角】FakeApp 教學

個po 開左兩日仲未有成品望兩眼 真係慢到咁？

第一步已經Fail左

cmd果步ffmpeg -i sceneA.mp4 -vf fps=30 "out%d.png"
直接出紅字sceneA.mp4: No such file or directory

你地全部自學？好勁咁

我都幾肯定一齊放左係users/user name

如果唔得
巴打直接上網搵個軟件 video to jpg~~~

dataA係影片；B係素材既話
左邊

未

巴打幾多秒彈一下?

可能係真係兩組圖好唔夾

我岩先一train就0.05 開始

一邊升 另一邊跌
搞咗一輪冇變化咁

10mins
因為我轉咗cpu

有冇人知memory error 點解決

圖少啲or圖的size少啲
再唔喺就改細setting啲數字

霸氣啲可以買多幾條ram

我16gb 啦

Using TensorFlow backend.
2018-01-29 22:23:29.915306: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu
\PY\35\tensorflow\core\platform\cpu_feature_guard.cc:137] Your CPU supports inst
ructions that this TensorFlow binary was not compiled to use: AVX AVX2
2018-01-29 22:23:30.403306: E C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu
\PY\35\tensorflow\stream_executor\cuda\cuda_driver.cc:406] failed call to cuInit
: CUDA_ERROR_NO_DEVICE
2018-01-29 22:23:30.968306: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu
\PY\35\tensorflow\stream_executor\cuda\cuda_diagnostics.cc:158] retrieving CUDA
diagnostic information for host: shigure-PC
2018-01-29 22:23:30.968306: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu
\PY\35\tensorflow\stream_executor\cuda\cuda_diagnostics.cc:165] hostname: shigur
e-PC
Focus the training preview window and press 'q' to stop training and save the mo
del
Traceback (most recent call last):
File "train.py", line 106, in <module>
File "training_data.py", line 17, in get_training_data
File "image_augmentation.py", line 21, in random_warp
AssertionError: MESSAGE: Your images are the wrong size. Images used for trainin
g should be 256x256 images produced by the Align tool.
[4864] Failed to execute script train

明明係256X256
但你地
extract出黎果d係jpeg定png

Png

都話用cpu係冇用 train一世都未得

果解我extract前素材係png但出完aligned既都係jpeg

但我fakeapp file type係png

train係train個model 你train完一次用其他人樣都唔使再train
google "deep learning"啦
如果唔係都唔會咁爆啦單野