CVPR 2023 to Tackle Bias Mitigation and Fair Representation in Artificial Intelligence

WHO: The Computer Vision and Pattern Recognition Conference (CVPR) is the preeminent computer vision event for new research in support of artificial intelligence (AI), machine learning (ML), augmented and virtual reality (AR/VR), deep learning, and much more.

WHAT: Given the growing discussion around ethical practices in AI, and the societal harm that bias and underrepresentation can generate, CVPR sessions speak to ways to mitigate bias and ensure fair representation.

From Model-Agnostic Gender Debiased Image Captioning, which outlines a framework for correcting gender misclassification and changing gender-stereotypical words to more neutral ones, to DeAR: Debiasing Vision-Language Models (VLMs) With Additive Residuals, which presents a novel model that can significantly improve the fairness of output representations of larger pre-trained VLMs, CVPR 2023 will explore this area of focus and the industry efforts being developed to address it.

Other papers focused on this topic include, but are not limited, to:

• BiasAdv: Bias-Adversarial Augmentation for Model Debiasing
  Authors: Jongin Lim, Youngdong Kim, Byungjai Kim, Chanho Ahn, Jinwoo Shin, Eunho Yang, Seungju Han
• Bias-Eliminating Augmentation Learning for Debiased Federated Learning
  Authors: Yuan-Yi Xu, Ci-Siang Lin, Yu-Chiang Frank Wang
• Bias in Pruned Vision Models: In-Depth Analysis and Countermeasures
  Authors: Eugenia Iofinova, Alexandra Peste, Dan Alistarh
• Bias Mimicking: A Simple Sampling Approach for Bias Mitigation
  Authors: Maan Qraitem, Kate Saenko, Bryan A. Plummer
• Uncurated Image-Text Datasets: Shedding Light on Demographic Bias
  Authors: Noa Garcia, Yusuke Hirota, Yankun Wu, Yuta Nakashima

WHEN/WHERE: CVPR 2023 takes place 18-22 June 2023, at the Vancouver Convention Center, Vancouver, Canada. For date and time specifics of the aforementioned sessions, and all other sessions at CVPR, please visit https://cvpr2023.thecvf.com/.

WHY: Sponsored by the IEEE Computer Society (CS) and the Computer Vision Foundation (CVF), CVPR 2023 delivers the important advances in all areas of computer vision and pattern recognition and the various fields and industries they impact.

With a record number of paper submissions, 9,155, an increase of 12 percent over last year, and an acceptance rate of just 25.8 percent, CVPR 2023 will deliver a program of the highest of quality that will enable attendees to discover, understand and gain critical knowledge around the computer vision and pattern recognition topics that are transforming the industry.

HOW: Interested industry professionals can register for CVPR 2023 by visiting https://cvpr2023.thecvf.com/Conferences/2023/Pricing2.

Members of the media can request a media pass by visiting https://cvpr2023.thecvf.com/Conferences/2023/MediaPass.

FOR MORE INFORMATION:
CVPR 2023 PR Team
media@computer.org