Comments (4)
Hi SlingHe, Thanks for your interest.
The shift-pooling is implemented in the function ``dynamic_local_filtering'' at https://github.com/dingmyu/D4LCN/blob/master/models/resnet_dilate.py#L13. Note that here we provide a simplified and faster version with comparable performance.
It can be implemented simply, for example:
y = torch.cat((x[:, -1:, :, :], x[:, :-1, :, :]), dim=1)
z = torch.cat((x[:, -2:, :, :], x[:, :-2, :, :]), dim=1)
x = (x + y + z)/3
If you modify the parameter
from d4lcn.
Feel free to reopen it if you have any further questions.
from d4lcn.
Hi, what do you mean a better result than paper may be achieved? Did you try different settings of
from d4lcn.
Hi, what do you mean a better result than paper may be achieved? Did you try different settings of
$n_f$ ? Can you please provide some experimental results or findings with different$n_f$ ? Thanks!
Hi Danny,
I did this experiment and found that
However, when I clean up this code, I found that simplifying the shift-pooling and modifying BN / relu / dropout sometimes lead to better results. For example,
Thank you.
from d4lcn.
Related Issues (20)
- Reproduce 3DNet+CL HOT 11
- Optimizer.step() & scheduler.step() in wrong order when restoring HOT 1
- Depth maps estimation
- Reproducing AP|R11 results
- Source of Randomness HOT 1
- data distribute
- Confuion in building project
- sh train.sh undefined symbol: __cudaRegisterFatBinaryEnd HOT 2
- Where is the function ‘projection_ray_trace’?
- Meaning of the Results HOT 2
- result HOT 2
- how to calculate mean and std HOT 1
- How can i train the results like you in paper?
- Replicate results from the paper HOT 5
- sh train.sh TypeError : cannot creat 'generator' instances
- ModuleNotFoundError: No module named 'lib.imdb_util'
- Depth Images for Testing set HOT 1
- Cannot replicate results from the paper
- Training my own dataset
- Question about "gpu_nms"
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 d4lcn.