Comments (6)
First question is how to upsample to generate the final output heatmap?
- Bilinear upsampling will give more accurate gradient back-propagation for each pixel. But in testing, directly upsampling cannot produce the heatmap of higher resolution, which probably reduce the gain. Similar experiment is done in https://github.com/chenyilun95/tf-cpn/issues/4, which may show it doesn't work with better gradient in high resolution.
- Skip-connection with the lower feature maps, but their semantics aren't clear probably.
- Deconv: recent work (Simple Baseline for Human Pose Estimation) says it's fine with deconvolution layer. But they still upsample the output to 64x48. If that works, it might works as well in higher resolution output.
Nevertheless, that's only my viewpoints. Experiment results says louder !
from tf-cpn.
I apologize for my ambiguous expression.
my question is that the NET 's last layer output is 64x48,which is(W/4,H/4).
how about change the last layer output to 256x192,which is (W,H).
so that orig-img (W,H)->(W/2,H/2)->(W/4,H/4)->.....->(W/4,H/4)->(W/2,H/2)->(W,H),(pre-heatmap)
pixel to pixel match between orig-img and pre-heatmap will increase AP ?
Thanks for your response .
from tf-cpn.
Excuse me... I'm now confused ... how do you change the last layer output to 256x192 ?
from tf-cpn.
for exmaple
1,add some intermediate layer(W/2,H/2) by (Bilinear upsampling / Deconv/Skip-connection )
2,and(Bilinear upsample / Deconv/Skip-connect) it to(W,H).
from tf-cpn.
emmmm... then I think the above comments are my response... Generally, I tend to think it won't work considering efficiency and effectiveness.
from tf-cpn.
@chenyilun95,Thank you,I get it.
I note that most people make the last layer output to 64* 64 (Hourglass Net etc.), 64*48(yours).
so the best practice of last layer output is (W/4,H/4)?
Thanks for your response ,I will close this issue.
from tf-cpn.
Related Issues (20)
- Has anyone tested on the MPII dataset?
- Strange results HOT 1
- About detect single person HOT 16
- About code running environment
- How do I test with pictures?
- bbox is necessary in Test? HOT 2
- An error occurred after I executed: `python3 mptest.py -d 0-1 -m log/model_dump/snapshot_350.ckpt`
- global_loss += tf.reduce_mean(tf.square(global_out - global_label)) / len(labels); global_loss /= 2;.so damn hard to understand
- why? gk15 = (23, 23) gk11 = (17, 17) gk9 = (13, 13) gk7 = (9, 9)
- strange results
- > @dagongji10 Excuse me, can I ask you a question? Must I provide bbox of pictures when I test a picture?Could you please give me a script which can visualize test results like what you show? Thank you very much HOT 3
- coco dataset? which year?
- train train
- "read none image" during training? HOT 1
- from nets.basemodel import resnet50, resnet_arg_scope, resnet_v1
- labels = [label15, label11, label9, label7]
- How to apply inference on own dataset without GT_bbox(.json)?
- How to decrease Batch size HOT 4
- Using just GlobalNet
- is it possible to add new joint in the skel and retrain the model
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 tf-cpn.