Comments (4)
GL-NF only uses Triplet Loss with Global distance. The difference to open-reid triplet loss example includes
- whether to use standard Market1501 train/test split, false in open-reid, but true here
- the type of data augmentation, mirror + cropping in open-reid, only mirror here
- batch size may be different, here 128 images
- whether to use an extra embedding layer (128-dim) after average pooling (2048-dim), used in open-reid, but here not
- whether to normalize the feature before computing triplet loss and for testing, done in GL-NF. This has little influence when triplet loss is only trained with global distance.
- the epoch at which the learning rate starts to decay, epoch 100 for open-reid and 75 here
Thanks :)
from alignedreid-re-production-pytorch.
Thanks for sharing your training details!
However I have different experiment results comparing with your fifth item. After feature normalization before computing triplet loss, the CMC score decreases ~8%. And I find that with or without extra embedding layer doesn't affect the accuracy. So I doubt whether normalization caused the different baseline result?
Best,
from alignedreid-re-production-pytorch.
- I found that for
Global Distance + Triplet Loss
, i.e. the commonly used triplet loss paradigm, normalizing feature or not has little influence. - Without normalization, I found it ok to train
Global Distance + Local Distance + Triplet Loss
. So later experiments only consider not normalizing features. - By
decreases ~8%
, do you meanGL-LL-NF-LHSFLD-TWLD
is worse thanGL-LL-NNF-LHSFLD-TWLD
by~8
CMC Rank-1 points? - BTW, the percentage of satisfying margin, i.e.
distance(anchor, positive) + margin < distance(anchor, negative)
is a good signal to check the status of convergence. It should be at least over90%
for both global and local distance. Do you find something abnormal in this signal at the end of training?
from alignedreid-re-production-pytorch.
With more experiments, I find that for vanilla triplet loss, with and without normalizing feature indeed makes difference, the later being better. Some results are in the provided Excel file.
Thank you for discussion.
from alignedreid-re-production-pytorch.
Related Issues (20)
- top-k结果可视化 HOT 2
- CUHK03和DUKE上的识别率 HOT 1
- How to inference my own test set
- Keys not found in source state_dict HOT 1
- Global Feature Extraction HOT 1
- Local feature dimensions
- About performance on market1501 for global learning and mutual learning
- Is it generalised
- TypeError: __init__() got an unexpected keyword argument 'log_dir'
- AssertionError HOT 3
- 请问论文中的Resnet50-Xception结构是不是没有实现? HOT 1
- 论文复现的参数问题
- how to use the test data to draw picture just like roc missrate cmc?
- How to use without GPU? HOT 2
- could you send me a partitions.pkl about market1501 HOT 1
- 为啥用你提供的weight测得market也只有88.78的top1呢 HOT 1
- how to infer some images or videos
- RuntimeError: cannot perform reduction function min on tensor with no elements because the operation does not have an identity
- Poor performnace when reproducing evaluation on market1501 HOT 2
- would you help me to fix this error?
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 alignedreid-re-production-pytorch.