Robust (Deep) learning framework against dirty labels and beyond

Amirmasoud Ghiassi, Taraneh Younesian, Zhilong Zhao, Robert Birke, Valerio Schiavoni, Lydia Y. Chen

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

7 Citations (Scopus)

Abstract

Data is generated with unprecedented speed, due to the flourishing of social media and open platforms. However, due to the lack of scrutinizing, both clean and dirty data are widely spreaded. For instance, there is a significant portion of images tagged with corrupted dirty class labels. Such dirty data sets are not only detrimental to the learning outcomes, e.g., misclassified images into the wrong classes, but also costly. It is pointed out that bad data can cost the U.S. up to a daunting 3 trillion dollars per year. In this paper, we address the following question: how prevailing (deep) machine learning models can be robustly trained given a non-negligible presence of corrupted labeled data. Dirty labels significantly increase the complexity of existing learning problems, as the ground truth of label's quality are not easily assessed. Here, we advocate to rigorously incorporate human experts into one learning framework where both artificial and human intelligence collaborate. To such an end, we combine three strategies to enhance the robustness for deep and regular machine learning algorithms, namely, (i) data filtering through additional quality model, (ii) data selection via actively learning from expert, and (iii) imitating expert's correction process. We demonstrate three strategies sequentially with examples and apply them on widely used benchmarks, such as CIFAR10 and CIFAR100. Our initial results show the effectiveness of the proposed strategies in combating dirty labels, e.g., the resulting classification can be up to 50% higher than the state-of-the-art AI-only solutions. Finally, we extend the discussion of robust learning from the trusted data to the trusted execution environment.

Original languageEnglish
Title of host publicationProceedings - 1st IEEE International Conference on Trust, Privacy and Security in Intelligent Systems and Applications, TPS-ISA 2019
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages236-244
Number of pages9
ISBN (Electronic)9781728167411
DOIs
Publication statusPublished - 2019
Event1st IEEE International Conference on Trust, Privacy and Security in Intelligent Systems and Applications, TPS-ISA 2019 - Los Angeles, United States
Duration: 12 Dec 201914 Dec 2019

Publication series

NameProceedings - 1st IEEE International Conference on Trust, Privacy and Security in Intelligent Systems and Applications, TPS-ISA 2019

Conference

Conference1st IEEE International Conference on Trust, Privacy and Security in Intelligent Systems and Applications, TPS-ISA 2019
Country/TerritoryUnited States
CityLos Angeles
Period12/12/1914/12/19

Keywords

  • Active learning
  • Adversarial learning
  • Data filtering
  • Deep neural networks
  • Dirty labels
  • Trusted execution

Fingerprint

Dive into the research topics of 'Robust (Deep) learning framework against dirty labels and beyond'. Together they form a unique fingerprint.

Cite this