MCPcopy Index your code
hub / github.com/TJUMMG/3DFP_FCGAN

github.com/TJUMMG/3DFP_FCGAN @main

Chat with this repo
repository ↗ · DeepWiki ↗ · + Follow
70 symbols 186 edges 9 files 19 documented · 27%
What it actually does AI analysis from the code graph — generated when you open this
loading…
README

3DFP_FCGAN

Implementation of paper "3DFP-FCGAN : Face Completion Generative Adversarial Network with 3D Facial Prior" [2021JVCIR].

Configuration Environment

python 3.7.0 + tensorflow 1.13.0 + neuralgym 0.0.1

Preparation

1. Dataset

For CelebA: download the cropped face images. For CelebA HQ: dowmload the face images and then run cropped_image.py to generate cropped face images.

2. pretrained model

Download the pretrained model from https://pan.baidu.com/s/1Ib1KQdcNLg9u1hUX1_e9CA [Passeword: vlcv]

For evaluation

Run test.py.

Acknowledgement

Thanks to previous open-sourced repo:

Generative image inpainting

Core symbols most depended-on inside this repo

gen_conv
called by 46
inpaint_ops.py
conv2d
called by 11
util/util.py
resBlock
called by 10
predictor.py
dis_conv
called by 8
inpaint_ops.py
conv2d
called by 7
util/util_bn.py
local_patch
called by 6
inpaint_ops.py
conv
called by 5
ops.py
spectral_norm
called by 4
ops.py

Shape

Function 50
Method 17
Class 3

Languages

Python100%

Modules by API surface

inpaint_ops.py20 symbols
ops.py16 symbols
Model.py11 symbols
predictor.py10 symbols
util/util_bn.py6 symbols
util/util.py6 symbols
cropped_image.py1 symbols

For agents

$ claude mcp add 3DFP_FCGAN \
  -- python -m otcore.mcp_server <graph>

⬇ download graph artifact