Comments (2)
After checking the definition of semihard triplets.I have the same suggestion as you.Have you confirmed this by experiments?
I found the original version may result in mode collapse,i.e.,the distance are nan
after a few epochs training.
from siamese-triplet.
@aikudexiaohai We need to use the same definition of loss as we use when computing gradients, not sure why you would lose the margin there.
In semihard example, by the definition of the loss it can't be negative. So I don't know what you're suggesting.
There may be many variations of semi-hard negatives, the definition I used here was from the FaceNet paper.
from siamese-triplet.
Related Issues (20)
- Citing Work
- how to classify
- Tripletloss finally returns to zero HOT 1
- #how could i reload the pre-trained
- #how could i reload the pre-trained parameters when I use siamese
- #how could i load the primary pre-trained params when I using the siamese module #
- utils.py:FunctionNegativeTripletSelector - 'anchor_positive' referenced before assignment when len(label_indices) < 2 HOT 6
- Implementation of Triplet loss on CASIA Web Face Dataset
- training
- References HOT 1
- BalancedBatchSampler: classes vs samples HOT 1
- How to use this siamese model for classification task?
- batch_szie
- batch_size
- datasets.py HOT 1
- ValueError when using BalancedBatchSampler
- I got this on cifar10 test set
- About generating all possible triplets using combinations() function HOT 1
- how tesT
- how should we use it to classify?
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 siamese-triplet.