ZiqiaoPeng / SyncTalk

[CVPR 2024] This is the official source for our paper "SyncTalk: The Devil is in the Synchronization for Talking Head Synthesis"
https://ziqiaopeng.github.io/synctalk/
Other
1.35k stars 161 forks source link

No faces were detected. #163

Open ahkimkoo opened 5 months ago

ahkimkoo commented 5 months ago
0% 0/82 [00:00<?, ?it/s]/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py:1078: RuntimeWarning: invalid value encountered in cast
  pred = (pred * 255).astype(np.uint8)
/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py:1084: RuntimeWarning: invalid value encountered in cast
  pred_depth = (pred_depth * 255).astype(np.uint8)
100% 82/82 [00:13<00:00,  6.24it/s]==> Finished Test.
100% 82/82 [00:14<00:00,  5.72it/s]
++> Evaluate at epoch 72 ...
  0% 0/82 [00:00<?, ?it/s]/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py:580: RuntimeWarning: invalid value encountered in cast
  inp = (inp * 255).astype(np.uint8)
/root/autodl-tmp/vms/synctalk/lib/python3.8/site-packages/face_alignment/api.py:147: UserWarning: No faces were detected.
  warnings.warn("No faces were detected.")
Traceback (most recent call last):
  File "main.py", line 211, in <module>
    trainer.evaluate(test_loader)
  File "/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py", line 1039, in evaluate
    self.evaluate_one_epoch(loader, name)
  File "/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py", line 1403, in evaluate_one_epoch
    metric.update(preds, truths)
  File "/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py", line 589, in update
    lms_pred = self.get_landmarks(preds)
  File "/root/autodl-tmp/SyncTalk/nerf_triplane/utils.py", line 560, in get_landmarks
    lms = self.predictor.get_landmarks(img)[-1]
TypeError: 'NoneType' object is not subscriptable
  0% 0/82 [00:01<?, ?it/s]
yyjjww commented 5 months ago

the same question,help! @ZiqiaoPeng

SuroshAhmadZobair commented 5 months ago

make sure your input video is 512x512 and the speaker is in the center.

yyjjww commented 5 months ago

meet the requirement,i sure,help me please!

@.***

 

------------------ 原始邮件 ------------------ 发件人: "ZiqiaoPeng/SyncTalk" @.>; 发送时间: 2024年6月19日(星期三) 晚上11:52 @.>; @.**@.>; 主题: Re: [ZiqiaoPeng/SyncTalk] No faces were detected. (Issue #163)

make sure your input video is 512x512 and the speaker is in the center.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>