Comments (8)
Take a look at SingleDataset
in https://github.com/baudm/MONet-pytorch/blob/master/data/single_dataset.py
You can use it as-is as long as your image transform requirement is covered by the get_transform()
function.
I created a new class, CLEVRDataset
, because the get_transform()
function can't achieve the specific image transformations that are needed for CLEVR.
from monet-pytorch.
Also, check out the options via ./train.py --help
from monet-pytorch.
Thanks a lot. I'm doing my datasets. Must the input images be 64x64? Can I change the input size of images?
from monet-pytorch.
The current implementation currently accepts 64x64 or 128x128. You can use 128x128 images by passing full_res=True
to ComponentVAE
.
If you want to use other image sizes, you need to specify the correct number of hidden units in the nn.Linear
layer of ComponentVAE
to accommodate the change in the feature map dimensions.
from monet-pytorch.
I forgot-- to use 128x128 images, you also need to uncomment the commented out AttentionBlock
s in the Attention
model. To use other image sizes, you need to fiddle with with the Attention
model because it contains nn.Linear
layers too.
from monet-pytorch.
OK. It's really a great work. And can you explain how to get the result of Figure 5 in your paper as you say "single latent z_k,i is traversed from -1.0 to +1.0" 3.2 in disentangled representations?
from monet-pytorch.
Sorry, but I'm not one of the authors. Better to contact the corresponding author indicated in the paper.
from monet-pytorch.
Ok. Thank you very much.
from monet-pytorch.
Related Issues (7)
- Can't pickle local object HOT 2
- Loss function of the monet HOT 2
- questions about using code base: Monet-pytorch HOT 2
- question about the scope HOT 1
- Something wrong with ./scripts/generate_monet_montage.sh after testing using pre-trained model HOT 2
- AttributeError: Can't pickle local object 'StringEncoder.<locals>.EncodeField' HOT 1
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 monet-pytorch.