The RMACC Women in HPC Chapter is pleased to host a two-part program that will explore problematic practices of datasets and AI - which can be prejudicial and biased - and the consequences.

Part one (December 16, 2020 from 3-4 PM MST) is an introductory presentation by Dr. Vinay Prabhu, who will provide a background on existing practices. The presentation by Dr. Prabhu will be based on his work, covered by VentureBeat, which ultimately led to the withdrawal of problematic tiny image datasets. (Part one video can be found here.)

Part two (January 20, 2021 from 3-4 PM MST) will include a panel discussion and an interactive session with attendees to collect best practices and actionable takeaways for the HPC community.

Large datasets: A Pyrrhic win for computer vision?

In this talk, we will investigate problematic practices and consequences of large-scale vision datasets. We examine broad issues such as the question of consent and justice as well as specific concerns such as the inclusion of verifiably pornographic images in machine learning datasets. We survey the landscape of harm and threats both society broadly and individuals face due to uncritical and ill-considered dataset curation practices. We then propose possible courses of correction and critique the pros and cons of these. To conclude, we provide details of the associated open-source project created in this endeavor that also hosts the meta-data sets generated for the community to build on.

About the Speaker - Dr. Vinay Prabhu

Vinay Prabhu is currently on a mission to model human kinematics using motion sensors on smartphones paving the way for breakthroughs in areas such as geriatric care, passive authentication, neuro-degenerative disease modeling, fitness and augmented reality. He is currently the Chief Scientist at UnifyID Inc and has peer-reviewed contributions spanning Physical layer wireless communications, Estimation theory, Information Theory, Network Sciences, and Machine Learning. His recent research projects include Deep Connectomics networks, Grassmannian initialization, SAT: Synthetic-seed-Augment-Transfer framework, and the Kannada MNIST dataset. He holds a PhD from Carnegie Mellon University and an MS from the Indian Institute of Technology-Madras. In his spare time, he works on his cricketing skills and generative art projects, some of which have made it to the playa at Black Rock City.

Part 2: Panel Discussion on January 20th

This moderated panel will include 25 minutes of introductory discussion followed by 35 minutes of wider discussion involving questions from the audience. We encourage attendees to participate in the discussion!

Our confirmed panelists are:

Vinay Prabhu, Chief Scientist at UnifyID Inc, PhD Carnegie Mellon
Alice Xiang, former Head of Fairness, Transparency, and Accountability Research at PartnershipAI and current senior research scientist at SonyAI
Matissa Hollister, a fellow at the World Economic Forum in ML/AI Platform from McGill University. She is an expert in organizational behavior and bias in AI
Daniel Lim, Senior Director of Experience Design at Salesforce and fellow at the World Economic Forum
Jeanne Holm, Deputy Mayor of LA, former City’s Chief Data Officer, and Mayor’s Senior Technology Advisor

For more information about the Rocky Mountain Advanced Computing Consortium(RMACC) or the RMACC Women in HPC Chapter, email rmacc@colorado.edu.

Zoom Coordinates

Join from PC, Mac, iOS or Android: https://cuboulder.zoom.us/j/693542827

Or join by phone:

+1 646 558 8656 (US Toll) or +1 408 638 0968 (US Toll)

Meeting ID: 693 542 827

(Photo credit: h heyerlein on Unsplash)