Comments (2)
This is because the image size is reduced each time it passes the layer.
At MNIST, I set nb_block = 2 so that the dense block and transition layer pass only two times.
(See ReadME)
The image size is halved from the transition layer (because, avg pooling)
It is also reduced by half in the first conv layer (7 * 7, stride = 2) (also max pooling too)
To summarize, the image size of MNIST is 28 * 28.
So,
28 -> 14(first conv) -> 6(max pool, The reason for not half is because of kernel size.)
-> 3 (dense +transition) -> 1 (dense+transition) -> error !
Therefore, it is necessary to set the number of dense blocks according to the input image.
Thank you
from densenet-tensorflow.
I modified the code so that it can be run on MNIST.
from densenet-tensorflow.
Related Issues (20)
- A bug in transition layer HOT 29
- An question
- About pretrained model for transfer learning
- Hellow,I'm a beginner of DL
- why throws a ResourceExhaustedError after a 30 epochs training process?
- growing filter
- Testing from a Saved checkpoint
- shuffling data
- does not converge
- how to standardize a single image in testing phase?
- freeze densnet ckpt to pb problems.
- The graph couldn't be sorted in topological order.
- The inputs of all 1*1 layers in a dense block include the output of the transition layer?
- Model compilation failing with ValueError: Variable dense_1_bottleN_0_batch1/beta already exists, disallowed HOT 1
- pretrained weight for accuracy verification
- weights for the models provided HOT 4
- A bug in dense_block HOT 1
- Why the axis of concat layer is 3 HOT 2
- Data preprocessing HOT 1
- Implementation without compression or bottleneck layers
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 densenet-tensorflow.