Coder Social home page Coder Social logo

tpn-1's Introduction

Transductive Propagation Network

Code for ICLR19 paper: Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learning. pdf

Requirements

  • Python 3.5
  • Tensorflow 1.3+
  • tqdm

Data Download (miniImagenet and tieredImagenet)

Please download the compressed tar files from: https://github.com/renmengye/few-shot-ssl-public

mkdir -p data/miniImagenet/data
tar -zxvf mini-imagenet.tar.gz
mv *.pkl data/miniImagenet/data

mkdir -p data/tieredImagenet/data
tar -xvf tiered-imagenet.tar
mv *.pkl data/tieredImagenet/data

TPN mini-5way1shot

python train.py --gpu=0 --n_way=5 --n_shot=1 --n_test_way=5 --n_test_shot=1 --lr=0.001 --step_size=10000 --dataset=mini --exp_name=mini_TPN_5w1s_5tw1ts_rn300_k20 --rn=300 --alpha=0.99 --k=20
python test.py --gpu=0 --n_way=5 --n_shot=1 --n_test_way=5 --n_test_shot=1 --lr=0.001 --step_size=10000 --dataset=mini --exp_name=mini_TPN_5w1s_5tw1ts_rn300_k20 --rn=300 --alpha=0.99 --k=20 --iters=81500

TPN mini-5way5shot

python train.py --gpu=0 --n_way=5 --n_shot=5 --n_test_way=5 --n_test_shot=5 --lr=0.001 --step_size=10000 --dataset=mini --exp_name=mini_TPN_5w5s_5tw5ts_rn300_k20 --rn=300 --alpha=0.99 --k=20
python test.py --gpu=0 --n_way=5 --n_shot=5 --n_test_way=5 --n_test_shot=5 --lr=0.001 --step_size=10000 --dataset=mini --exp_name=mini_TPN_5w5s_5tw5ts_rn300_k20 --rn=300 --alpha=0.99 --k=20 --iters=50100

TPN tiered-5way1shot

python train.py --gpu=0 --n_way=5 --n_shot=1 --n_test_way=5 --n_test_shot=1 --lr=0.001 --step_size=25000 --dataset=tiered --exp_name=tiered_TPN_5w1s_5tw1ts_rn300_k20 --rn=300 --alpha=0.99 --k=20

TPN tiered-5way5shot

python train.py --gpu=0 --n_way=5 --n_shot=5 --n_test_way=5 --n_test_shot=5 --lr=0.001 --step_size=25000 --dataset=tiered --exp_name=tiered_TPN_5w5s_5tw5ts_rn300_k20 --rn=300 --alpha=0.99 --k=20

Citation

If you use our code, please consider cite the following:

  • Yanbin Liu, Juho Lee, Minseop Park, Saehoon Kim, Eunho Yang, Sungju Hwang, Yi Yang. Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learning. In Proceedings of 7th International Conference on Learning Representations (ICLR), 2019.

@inproceedings{liu2019fewTPN,
	title={Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learning},
	author={Yanbin Liu and 
		Juho Lee and 
		Minseop Park and 
		Saehoon Kim and 
		Eunho Yang and 
		Sungju Hwang and 
		Yi Yang},
booktitle={International Conference on Learning Representations},
year={2019},
}

tpn-1's People

Contributors

csyanbin avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

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

  • Vue.js photo Vue.js

    ๐Ÿ–– 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. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

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 โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.