Comments (1)
I guess you mean how the model differentiates the search region from the targets?
For mixformer architecture with convolution implicitly encoding the positional embeddings, it can learn the positional differentiations between the two ones.
For MixViT with substituting the backbone with ViT, we have experimented three different kinds of positional embeddings for search and targets respectively and obtained similar performance. It indicates that additional pos. emb. for differentiating the two ones are unnecessary.
More details can be found at the journal version of MixFormer (will be released soon at Arxiv).
from mixformer.
Related Issues (20)
- Where are the CVT models? HOT 1
- Stage 2 training HOT 7
- Multiple static targets HOT 4
- MixViT CovMAE HOT 1
- Model weights of MixViT-L HOT 4
- test on my own dataset HOT 2
- Could you please open-source the TrackMae pre-training code as soon as possible?
- AttributeError: 'MixFormerOnline' object has no attribute 'online_sizes' HOT 1
- Process multiple templates in a batch HOT 1
- MixFormer attention question (please...) HOT 1
- Mixformer ONNX export tool
- Score prediction module
- the cuda version cudatoolkit=10.2
- Performance on GOT-10k HOT 4
- Use pretrained model on CPU HOT 2
- from test.evaluation.environment import EnvSettings ModuleNotFoundError: No module named 'test.evaluation'; 'test' is not a package
- Can we run it on windows wiht torch 2.0.1 ? HOT 1
- 有关指标计算
- About training on your own dataset
- Test on our own dataset with multiple instances per video HOT 3
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 mixformer.