Open DylanMaeng opened 5 years ago
@DylanMaeng Hi, I meet this problem too. Nothing changes.. Have you solved that?
me too
@DylanMaeng Hi, I meet this problem too. Nothing changes.. Have you solved that?
Sorry,i don't.
options = {
"use_smoothed_bbox": True,
"use_kalman_filter": True,
"use_auto_downscaling": False,
"bbox_moving_avg_coef": 0.65,
"min_face_area": 35 * 35,
"IMAGE_SHAPE": model.IMAGE_SHAPE,
# ===== Tunable =====
"kf_noise_coef": 1e-3,
"use_color_correction": "hist_match",
"detec_threshold": 0.8,
"roi_coverage": 0.9,
"enhance": 0.,
"output_type": 3,
"direction": "AtoB", # ==================== This line determines the transform direction ====================
} in this cell, swith your output_type to 1
@pmdw620, Many thanks... Nothing changed for me, after changing output_type to 1, May I request for your kind help ? Thanks again
You should change it before you run the cell. On Fri, Mar 13, 2020, 23:01 ajsirsik notifications@github.com wrote:
@pmdw620 https://github.com/pmdw620, Many thanks... Nothing changed for me, after changing output_type to 1, May I request for your kind help ? Thanks again
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/shaoanlu/faceswap-GAN/issues/122#issuecomment-598687260, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKLSZL4O4P5O3JVIGOJBV5DRHIOBRANCNFSM4GO2PMDA .
Many thanks @pmdw620 for quick reply and extending helping hand, I changed it before i run the cell, getting results unlike shown in image posted by DylanMaeng earlier. but face in output video did not get swap with another face. Do I need to execute something more than colab demo ?
These are results after 30k iterations. Am I missing something here ?
Maybe because you have not trained your models enough, I've take a look at your notebook, the recent training stops at 600
On Sat, 14 Mar 2020 at 00:33, ajsirsik notifications@github.com wrote:
Here is my colab notebook:
https://colab.research.google.com/drive/1icOuUA07Ra0-siMwZej9EF9Bhng-iDZj
On Fri, Mar 13, 2020 at 7:02 PM Ajinkya Sirsikar < theajinkya.sirsikar@gmail.com> wrote:
Dear sir,
Many thanks for your reply and kind help. I am new to this field and still learning. Your article and code is a great help for newbies like me.
I changed it before i run the cell, getting results unlike shown in image posted by DylanMaeng earlier. but face in output video did not get swap with another face. May be, I must be missing something, Do I need to execute something more than colab demo.
Sharing my colab notebook, just for your reference.
Thanks again for extending helping hand
Regards Ajinkya
On Fri, Mar 13, 2020 at 5:51 PM pmdw620 notifications@github.com wrote:
You should change it before you run the cell. On Fri, Mar 13, 2020, 23:01 ajsirsik notifications@github.com wrote:
@pmdw620 https://github.com/pmdw620, Many thanks... Nothing changed for me, after changing output_type to 1, May I request for your kind help ? Thanks again
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <
https://github.com/shaoanlu/faceswap-GAN/issues/122#issuecomment-598687260
, or unsubscribe <
https://github.com/notifications/unsubscribe-auth/AKLSZL4O4P5O3JVIGOJBV5DRHIOBRANCNFSM4GO2PMDA
.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/shaoanlu/faceswap-GAN/issues/122#issuecomment-598694498 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AM3PBYNS6M45DQSXIERQMB3RHIQNXANCNFSM4GO2PMDA
.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/shaoanlu/faceswap-GAN/issues/122#issuecomment-598722111, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKLSZL4Y3C23LSTU5PWOQELRHIY4JANCNFSM4GO2PMDA .
Hi,shaoanlu! You are very cool! I try to train the faceswap-GAN_colab_demo.ipynb.It takes about 6 hours.All is well.Preview is OK. But the output.mp4 is as follow: I would appreciate it if you could tell me what I need to do.