A Hands-on Tutorial for
Learning with Noisy Labels
Date / Time: August 20th, 2023 (AM)
A brief intro of our tutorial
Learning with noisy labels is a pervasive challenge in different research and application areas that require supervised training data. This tutorial provides hands-on instructions on how to implement the existing and most relevant solutions using Jupyter Notebook examples.
Outline The tutorial will be given primarily using Jupyter Notebook examples with proper explanations, pointers, and discussions.
Module 1: Setups This module will set up the basic settings and provide examples that show how one can simulate noisy labels for controlled experiments.
Module 2: Learning the noise rate in the labels without knowing the ground truth This module will go through examples showing how one can estimate the hidden noise transition matrix that controls the generation of noisy labels, without using ground truth annotations. The knowledge of noise rate plays a central role in understanding the quality of data and in building robust training approaches.
Module 3: Learning algorithms that handle noisy labels This module will provide examples showing how one would implement a learning algorithm specifically designed for handling noisy labels.
Module 4: Noise label detection This section will go through examples that explain how one can implement a detection algorithm to identify the wrong labels in a training dataset.
Module 5: Real-world datasets This module will go through examples using real-world datasets that contain noisy human annotations. We will primarily focus on using the CIFAR-N dataset.
For module details, please refer to the Schedule page.
The Speakers
Yang Liu
Assistant Professor
at UC Santa Cruz
Zhaowei Zhu
PhD Candidate
at UC Santa Cruz
Jiaheng Wei
PhD Candidate
at UC Santa Cruz
Hao Cheng
PhD Student
at UC Santa Cruz
For details information of presenters, please refer to the Presenters page.