Giter Club home page Giter Club logo

transformer-uvos's Introduction

Transformer-UVOS

This is a repo offers a strong baseline for Unsupervised Video Object Segmentation(UVOS) using Transformer

Accuracy

With Res50 as backbone and image size (640x352) during training, using only single-scale for testing On DAVIS 2016(val):

Method J_mean J_recall J_decay F_mean F_recall F_decay
Ours(Res50-FPN, 640x352) 0.777 0.915 0.066 0.766 0.859 0.043
Anchor-Diffusion(Res101-Deeplabv3, 854x480) 0.782 --- --- 0.771 --- ---

Architecture

ResNet50-FPN + Transformer + Simple Decoder

Training

You need to first download the DAVIS dataset

The network is initialized using ResNet50-FPN pre-trained on COCO dataset, which can be downloaded from here.

cd model/transformer/davis.transformer.fpn.R50.random_sample/
sh run.sh

Acknowledgement

Idea inspired by Anchor-Diffusion.

Codes based on TorchSeg, Detectron2 and Detr

Contact

Feel free to contact me if you have any questions : [email protected]

transformer-uvos's People

Contributors

dulucas avatar

Stargazers

avatar avatar avatar avatar avatar avatar avatar avatar

Watchers

avatar

Forkers

ym24lbj yahoo0742 huayangcheng

transformer-uvos's Issues

How to generate “img_pair_list_dense.txt”？

Interesting idea and thanks for sharing the code. I was wandering how the each sample formed "reference image and current image". And would you like to share the "img_pair_list_dense.txt"? Really appreciate.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.