Comments (4)
Due to several downsampling and upsampling, the size of the heatmaps can indeed change. I think that the original dimensions must be divisible by 8 to keep the same resolution as in the output.
However, if I remember correctly, the output size will always be bigger than the input size, and you can simply crop the output to make it fit with the input one, removing the right and bottom borders. In your case, you could do this:
h, w = image_tensor.shape[2:4]
junction_heatmap_cropped = junction_heatmap[:, :h, :w]
line_heatmap_cropped = line_heatmap[:, :h, :w]
from kornia.
@ducha-aiki any insights here ?
from kornia.
@ogencoglu Actually, you are not supposed to use any of the heatmaps, the "official" output is line segments and descriptors :)
Anyway, let's tag authors here @rpautrat
from kornia.
@ducha-aiki I know. But I need and am using heatmaps for other purposes as well.
I confirm that dims need to be divisible by 8 to have the same resolution. Thanks. This is enough I think.
from kornia.
Related Issues (20)
- Consolidate model architectures in separate file
- RanAug test failing
- [Bug] RandomJPEG does fails if sides are not divisible by 16 HOT 3
- keepdim not working as expected in AugmentationSequential HOT 1
- `AugmentationSequential` does not support instance masks shape (N, H, W) HOT 1
- Improve Image Matching docs HOT 2
- `RandomJPEG` can't handle input on different from CPU device. HOT 5
- Bug in normalize_min_max HOT 2
- Update apply_colormap HOT 5
- geometry.transform.build_pyramid max_level definition HOT 1
- cv2.edgePreservingFilter this filter support?
- RANSAC.max_samples_by_conf returns negative numbers HOT 1
- Cannot compile `torch.jit.script(kornia.geometry.warp_perspective)` because of `KORNIA_CHECK_IS_TENSOR` HOT 2
- Move WunschLineMatcher to kornia.feature.matching Module HOT 2
- find_essential and run_5point are not working correctly with batch size>1 HOT 5
- Update pytorch to 2.3.0 on CI
- Refactor `SOLD2Net` to Support Dataclasses for Configuration HOT 1
- AttributeError: 'list' object has no attribute 'ndim' with RandomTransplantation and batch with keys HOT 2
- Including Steerers HOT 2
- mypy error: Argument 4 to "AugmentationSequential" has incompatible type "RandomTransplantation"; expected "_AugmentationBase | ImageSequential"
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 kornia.