IJCAI Tutorial T-22

A Hands-on Tutorial for Learning with Noisy Labels

T22: A Hands-on Tutorial for Learning with Noisy Labels

Presnetors: Yang Liu; Zhaowei Zhu; Jiaheng Wei; Hao Cheng

Time: Aug 20th (Morning)

Materials:

Sec1_Intro.ipynb: Setups of Noisy Labels

Sec2_Est_Detect.ipynb: Includes Module 1: Estimate Transition Matrix; Module 2: Find label errors.

Sec3_lnl_cifarn.ipyn: Includes Module 3: Robust learning with noisy labels; Module 4: Real-world label noise and performance;

Sec4_challenge_future.ipynb: Includes Module 5: Challenges and Future Directions.

To run the code with Docta.ai, please install docta.ai first (and star if you think it is helpful).

pip install docta.ai

Name		Name	Last commit message	Last commit date
Latest commit History 90 Commits
cifarn/data		cifarn/data
config		config
data		data
models		models
noise_label/cifar-10		noise_label/cifar-10
tutorial_imgs		tutorial_imgs
tutorials.old		tutorials.old
.gitignore		.gitignore
README.md		README.md
Sec1_Intro.ipynb		Sec1_Intro.ipynb
Sec2_Est_Detect.ipynb		Sec2_Est_Detect.ipynb
Sec3_lnl_cifarn.ipynb		Sec3_lnl_cifarn.ipynb
Sec4_challenge_future.ipynb		Sec4_challenge_future.ipynb
loss.py		loss.py
main.py		main.py
main_cifarn.py		main_cifarn.py
requirements.txt		requirements.txt
test.py		test.py