Comments (5)
Thanks for pointing this out. We will test it following your suggestion.
from idinvert.
Both the generator and encoder were trained using FFHQ datasets, which have already been aligned. Thus, if you want to invert the image from the Web, make sure that you have aligned the image first.
from idinvert.
Thanks for your answer, I will align the image before feeding into network.
Ps: In the perceptual model, did you try to calculate l2 loss in multi layer, instead of layer 22 only ( I calculate l2 loss in multi layer and notice that the result is much better )
from idinvert.
We don't try it on multi-layers. Better reconstruction? or better manipulation? or something else?
from idinvert.
Better Reconstruction, I notice that the mse loss between two image is much smaller when doing perceptual loss in multi-layers, and the reconstruct image look more "realistic". Here is some reconstruct image with and without multi-layer perceptual. ( with 1000 iteration)
Original Image:
Reconstruct with multi-layer perceptual loss:
Reconstruct with original perceptual loss:
Ps : I have try to algin and the result is the same (Like what you see in example above), I currently try to expand your work with styleganv2, will inform you when the work is finish.
from idinvert.
Related Issues (20)
- Error occurs when running invert.py HOT 1
- How to set gpu id in train.py HOT 2
- ModuleNotFoundError: No module named 'tensorflow.contrib.nccl' HOT 1
- About resolution in styleGAN HOT 2
- about the W space HOT 1
- why face with eyeglasses when aged? HOT 2
- Training an encoder using pretrained StyleGAN HOT 2
- CUDA, cuDNN and NCCL versions HOT 2
- Training problems HOT 1
- My own data set HOT 1
- Could you provide me with the code for calculating SWD?
- Will you update the training code for pytorch soon? HOT 2
- How many iterations are needed to generate similar faces? HOT 1
- Error in Inversion Task HOT 2
- tensorflow.python.framework.errors_impl.InternalError: Blas GEMM launch failed HOT 3
- Using Semantic Diffusion in other domains (that are not faces) HOT 1
- Encoder is not trained at all. HOT 1
- Is it correct to use interfaceGAN when dealing with the tower dataset? HOT 1
- All of the losses except the reconstruction loss don't change. HOT 6
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from idinvert.