Artificial Intelligence with
Biased or Scarce Data

(AIBSD)

In Conjunction with 36th AAAI Conference on Artificial Intelligence 2022

February 28, 2022 (Virtual Workshop Held on the Virtual Chair Blue Building Plenary Room (Direction))

About AIBSD 2022

About AIBSD 2022

With the increasing appetite for data in data-driven methods, the issues of biased and scarce data have become a major bottleneck in developing generalizable and scalable artificial intelligence solutions, as well as effective deployment of these solutions in real-world scenarios. To tackle these challenges, researchers from both academia and industry must collaborate and make progress in fundamental research and applied technologies. The organizing committee and keynote speakers of AIBSD 2022 consist of experts from both academia and industry with rich experiences in designing and developing robust artificial intelligence algorithms and tranferring them to real-world solutions. AIBSD 2022 provides a focused venue to discuss and disseminate research related to bias and scarcity topics in artificial intelligence.

Where

Virtual Workshop Held on the Virtual Chair Blue Building Plenary Room (Direction)

When

09:00 AM - 04:50 PM PST
Monday, February 28, 2022

Keynote Speakers

Rama Chellappa

Rama Chellappa

Johns Hopkins University

David Crandall

David Crandall

Indiana University

Bernt Schiele

Bernt Schiele

Max Planck Institute for Informatics

AIBSD 2022 Schedule (in PST) [pdf]

Kuan-Chuan Peng

Opening Remarks Kuan-Chuan Peng

Bernt Schiele

Keynote Bernt Schiele

Addressing Imbalance Problems, Robustness and Interpretability of Deep Learning in Computer Vision

Paper Presentation Jyoti Narwariya, Chetan Kumar Verma, Pankaj Malhotra, Lovekesh Vig, Easwar Subramanian, Sanjay P. P. Bhat

Electricity Consumption Forecasting for Out-of-distribution Time-of-Use Tariffs

Paper Presentation Ahmed Frikha, Denis Krompass, Volker Tresp

COLUMBUS: Automated Discovery of New Multi-Level Features for Domain Generalization via Knowledge Corruption

Paper Presentation Nils Rethmeier, Isabelle Augenstein

Long-Tail Zero and Few-Shot Learning via Contrastive Pretraining on and for Small Data

Paper Presentation Styliani Katsarou, Borja Rodríguez Gálvez, Jesse Shanahan

Measuring Gender Bias in Contextualized Embeddings

Break

Paper Presentation Megan Frisella, Pooya Khorrami, Jason Matterer, Kendra Kratkiewicz, Pedro Torres-Carrasquillo

Quantifying Bias in a Face Verification System

Paper Presentation Maliha Arif, Calvin P. Yong, Abhijit Mahalanobis

Background Invariant Classification on Infrared Imagery by Data Efficient Training and Reducing Bias in CNNs

Paper Presentation Xiaoyan Zhuo, Wolfgang Rahfeldt, Xiaoqian Zhang, Ted Doros, Seung Woo Son

DAP-SDD: Distribution-Aware Pseudo Labeling for Small Defect Detection

Break

Rama Chellappa

Keynote Rama Chellappa

Bias Mitigation in AI Systems

Paper Presentation Daniel Y. Fu, Mayee Chen, Michael Zhang, Kayvon Fatahalian, Christopher Re

The Details Matter: Preventing Class Collapse in Supervised Contrastive Learning

Paper Presentation Jinfeng Li, Nikita Bhutani, Alexander Whedon, Chieh-Yang Huang, Estevam Hruschka, Yoshihiko Suhara

Extracting Salient Facts from Company Reviews with Scarce Labels

Paper Presentation SangEun Lee, Soyoung Oh, Minji Kim, Eunil Park

Measuring Embedded Human-like Biases in Face Recognition Models

Paper Presentation Shunsuke Kogure, Kai Watabe, Ryosuke Yamada, Yoshimitsu Aoki, Akio Nakamura, Hirokatsu Kataoka

Age Shouldn't Matter: Toward More Accurate Pedestrian Detection via Self-Training

David Crandall

Keynote David Crandall

Training Data Bias, through the Eyes of a Child

Break

Paper Presentation Kumpei Ikuta, Hitoshi Iyatomi, Kenichi Oishi

Super-resolution for Brain MR Images from Significantly Small Amount of Training Data

Paper Presentation Qian Ren, Jie Chen

Dual Complementary Prototype Learning for Few-shot Segmentation

Paper Presentation Yilu Guo, Shicai Yang, Weijie Chen, Liang Ma, Di Xie, Shiliang Pu

Attract-and-Repulse: Towards Data Deficient Learning

Paper Presentation Sung-Feng Huang, Chyi-Jiunn Lin, Hung-Yi Lee

Meta-TTS: Meta-Learning for Few-Shot Speaker Adaptive Text-to-Speech

Ziyan Wu

Closing Remarks Ziyan Wu



Submission

Submission Instructions

We welcome full paper submissions (up to 8 pages, excluding references or supplementary materials). Please submit at the
AIBSD 2022 @ AAAI2022 CMT website

The paper submissions must be in pdf format and use the AAAI official templates. All submissions must be anonymous and conform to AAAI standards for double-blind review. The accepted papers will be posted on the workshop website and will not appear in the AAAI proceedings. At least one author of each accepted submission must present the paper at the workshop.  

Submission deadline: November 12, 2021 AOE Time  
Notification to authors: December 3, 2021 AOE Time
Camera ready deadline:  December 15, 2021 AOE Time

We invite the submission of original and high-quality research papers in the topics related to biased or scarce data. Accepted work will be presented as either an oral or a spotlight presentation. 


Topics

The topics for AIBSD 2022 include, but are not limited to:

  • Algorithms and theories for explainable and interpretable AI models
  • Application-specific designs for explainable AI, e.g., healthcare, autonomous driving, etc.
  • Algorithms and theories for learning AI models under bias and scarcity.
  • Performance characterization of AI algorithms and systems under bias and scarcity
  • Algorithms for secure and privacy-aware machine learning for AI
  • Algorithms and theories for trustworthy AI models
  • The role of adjacent fields of study (e.g, computational social science) in mitigating issues of bias and trust in AI
  • Continuous refinement of AI models using active/online learning
  • Meta-learning models from various existing task-specific AI models
  • Brave new ideas to learn AI models under bias and scarcity


Instructions of Publication Preparation

We plan to publish the accepted papers with Computer Sciences & Mathematics Forum by MDPI. For the authors of the accpeted papers, please format your paper using the official MDPI template (Microsoft Word or LaTeX). If you use the Microsoft Word template, please submit the editable .doc/.docx file via CMT. If you use the LaTeX template, please also submit all the source files (compressed as a .zip file) together with the compiled .pdf file via CMT. If the camera ready paper is not submitted in the correct format on time, then its publication cannot be guaranteed.

Camera ready deadline: March 15, 2022 AOE Time  

  • COLUMBUS: Automated Discovery of New Multi-Level Features for Domain Generalization via Knowledge Corruption.
    Ahmed Frikha, Denis Krompass, Volker Tresp.
    [paper]

  • Quantifying Bias in a Face Verification System.
    Megan Frisella, Pooya Khorrami, Jason Matterer, Kendra Kratkiewicz, Pedro Torres-Carrasquillo.
    [paper]

  • Background Invariant Classification on Infrared Imagery by Data Efficient Training and Reducing Bias in CNNs.
    Maliha Arif, Calvin P. Yong, Abhijit Mahalanobis.
    [paper]

  • Dual Complementary Prototype Learning for Few-shot Segmentation.
    Qian Ren, Jie Chen.
    [paper]

  • Measuring Embedded Human-like Biases in Face Recognition Models.
    SangEun Lee, Soyoung Oh, Minji Kim, Eunil Park.
    [paper]

  • Age Shouldn't Matter: Toward More Accurate Pedestrian Detection via Self-Training.
    Shunsuke Kogure, Kai Watabe, Ryosuke Yamada, Yoshimitsu Aoki, Akio Nakamura, Hirokatsu Kataoka.
    [paper]

  • The Details Matter: Preventing Class Collapse in Supervised Contrastive Learning.
    Daniel Y. Fu, Mayee Chen, Michael Zhang, Kayvon Fatahalian, Christopher Re.
    [paper] (Best Paper Award Winner)

  • Super-resolution for Brain MR Images from Significantly Small Amount of Training Data.
    Kumpei Ikuta, Hitoshi Iyatomi, Kenichi Oishi.
    [paper]

  • Long-Tail Zero and Few-Shot Learning via Contrastive Pretraining on and for Small Data.
    Nils Rethmeier, Isabelle Augenstein.
    [paper]

  • Measuring Gender Bias in Contextualized Embeddings.
    Styliani Katsarou, Borja Rodríguez Gálvez, Jesse Shanahan.
    [paper]

  • DAP-SDD: Distribution-Aware Pseudo Labeling for Small Defect Detection.
    Xiaoyan Zhuo, Wolfgang Rahfeldt, Xiaoqian Zhang, Ted Doros, Seung Woo Son.
    [paper]

  • Extracting Salient Facts from Company Reviews with Scarce Labels.
    Jinfeng Li, Nikita Bhutani, Alexander Whedon, Chieh-Yang Huang, Estevam Hruschka, Yoshihiko Suhara.
    [paper]

  • Attract-and-Repulse: Towards Data Deficient Learning.
    Yilu Guo, Shicai Yang, Weijie Chen, Liang Ma, Di Xie, Shiliang Pu.
    [paper]

  • Meta-TTS: Meta-Learning for Few-Shot Speaker Adaptive Text-to-Speech.
    Sung-Feng Huang, Chyi-Jiunn Lin, Hung-Yi Lee.
    [paper]

  • Electricity Consumption Forecasting for Out-of-distribution Time-of-Use Tariffs.
    Jyoti Narwariya, Chetan Kumar Verma, Pankaj Malhotra, Lovekesh Vig, Easwar Subramanian, Sanjay P. P. Bhat.
    [paper]

AIBSD 2022 Venue

venue

Virtual Workshop

AIBSD 2022 will be held virtually on the Virtual Chair Blue Building Plenary Room (Direction)) during 09:00 AM - 04:50 PM PST on Monday, February 28, 2022.

Venue Direction

Please enter the Virtual Chair platform and find the Blue Building Plenary Room as the following screenshots

Sponsors

Algorithms (ISSN 1999-4893; CODEN: ALGOCH) is a peer-reviewed open-access journal indexed in Emerging Sources Citation Index - ESCI (Web of Science), EI, Scopus (CiteScore 2.9), and MathSciNet. The European Society for Fuzzy Logic and Technology (EUSFLAT) is affiliated with Algorithms.

Data (ISSN 2306-5729) is an international, peer-reviewed, open access journal on data in science, with the aim of enhancing data transparency and reusability. The journal is now indexed in Emerging Sources Citation Index - ESCI (Web of Science) and Scopus (CiteScore 3.5), ranking Q2 in "Information Systems and Management."

Organizers

Kuan-Chuan Peng

Kuan-Chuan Peng

Mitsubishi Electric Research Laboratories (MERL)

Ziyan Wu

Ziyan Wu

UII America, Inc.



Program Committee

Abhishek Aich
Yunhao Ge
Georgios Georgakis
Xuan Gong
Yunye Gong
Hengtao Guo
Lipeng Ke
Kunpeng Li
Runze Li
Qin Liu
Yao Xiao
Zhi Xu
Fan Yang
Meng Zheng