Eric Slyman (they/he)

AI/CS Ph.D Student, Oregon State University

me.jpeg

Hi Folks! I am a Ph.D. student at the intersection of multimodal AI, human-computer interaction, and fairness in the Artificial Intelligence and Computer Science programs at Oregon State University, where I’m advised by Stefan Lee (and Minsuk Kahng before his relocation to Google).

My research evaluates large-scale vision-language models like CLIP, ViLBERT, LLaVA, and DALL-E. I enjoy auditing the real-world steps to get these models in production and determining mitigations to help promote fair outcomes. Frequently, this leads to human-centered methods that consider the entire lifecycle of a model: from data collection to deployment.

In addition to research, I am also the Co-President of OSU’s AI Graduate Student Association, where I organize social and professional events for our AI community and run the AI Application Support Program to provide application advising assistance to underrepresented people applying to graduate school.

news

Feb 27, 2024 Our paper FairDeDup: Detecting and Mitigating Vision-Language Fairness Disparities in Semantic Dataset Deduplication was accepted at CVPR 2024.
Feb 7, 2024 My research on AI Fairness was selected as a Featured Program for the 2024 State of Diversity at Oregon State.
Nov 11, 2023 Our paper “VALET: Vision-And-LanguagE Testing with Reusable Components” was accepted at the NeurIPS 2023 Queer in AI Workshop.
Nov 1, 2023 My work was featured in the article “Adressing Bias in AI” in the magazine Taking Action from the OSU Office of Insitituional Diversity.
Sep 25, 2023 I was invited to share my work on VLSlice with the People+AI Research Group (PAIR) at Google.

selected publications

  1. fairdedup.png
    FairDeDup: Detecting and Mitigating Vision-Language Fairness Disparities in Semantic Dataset Deduplication
    Eric SlymanStefan LeeScott Cohen, and Kushal Kafle
    Computer Vision and Patern Recognition, Jun 2024
  2. valet.png
    VALET: Vision-And-LanguagE Testing with Reusable Components
    Eric SlymanKushal Kafle, and Scott Cohen
    NeurIPS Queer in AI Workshop, Dec 2023
  3. vlslice.png
    VLSlice: Interactive Vision-and-Language Slice Discovery
    Eric SlymanMinsuk Kahng, and Stefan Lee
    International Conference on Computer Vision, Oct 2023