pikachupichu25/live-portrait-image 🖼️📝✓ → 🖼️

▶️ 149.3K runs 📅 Sep 2024 ⚙️ Cog 0.9.21
expression-transfer face-reenactment image-editing image-to-image

About

Match facial expression using a driving image using LivePortrait as a base

Example Output

Output

Example output

Performance Metrics

7.34s Prediction Time
202.82s Total Time
All Input Parameters 
{
  "source_image": "https://replicate.delivery/pbxt/LcFm23sapw1R5mAHAwNsM0I9HZStryqRkPqKn9QlEfBuhGnM/d38.jpg",
  "driving_image": "https://replicate.delivery/pbxt/LcFm2KjcJKx2kdlimwU9XvICytJIM0xwMmMe89jIsJmNBkAb/d9.jpg",
  "transfer_pose": "all",
  "post_processing_eye": false,
  "post_processing_lip": true
}
Input Parameters
source_image (required) Type: string
Source image
driving_image (required) Type: string
Driving image
transfer_pose Type: stringDefault: all
post_processing_eye Type: booleanDefault: false
post_processing_lip Type: booleanDefault: true
Output Schema

Output

Type: string • Format: uri

Example Execution Logs 
source_lip_ratio [[0.7220591]] target_lip_ratio [[0.7704249]]
/src/src/live_portrait_wrapper.py:331: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:274.)
c_d_lip_i_tensor = torch.Tensor([c_d_lip_i[0]]).to(self.device).reshape(1, 1) # 1x1
Version Details
Version ID
f60012e7f8ebeeeff2fccbe6230161be65da2a4a017490b9b9c556b3e8183fec
Version Created
September 13, 2024
Run on Replicate →