Wei-Lin (Kimberly) Hsiao 蕭瑋琳

Kimberly Hsiao

I am a PhD candidate in the Computer Science Department at the University of Texas at Austin, advised by Professor Kristen Grauman. My research interests are in computer vision, with a focus on computational models for fashion understanding. I develop techniques that recognize fine-grained details in clothing, captures interactions between fashion items and their human wearers, and generalizes to fashion's evolution with society. I have also been a Visiting Researcher at Facebook AI Research since March, 2020.

Email: kimhsiao at cs dot utexas dot edu / [CV]


Nov. 2020: I'm selected as one of the Rising Stars in EECS, 2020.

Sep. 2020: We've released the Diverse Body Shape Dataset: Link.

Mar. 2020: I'm very happy to join Facebook AI Research (FAIR) as a Visiting Researcher.

Feb. 2020: ViBE has been accepted as an Oral paper at CVPR 2020.


From Culture to Clothing: Discovering the World Events Behind A Century of Fashion Images
Wei-Lin Hsiao, Kristen Grauman
arXiv, 2021

Learning patterns of tourist movement and photography from geotagged photos at archaeological heritage sites in Cuzco, Peru
Nicole D. Payntar, Wei-Lin Hsiao, R. Alan Covey, Kristen Grauman
Journal of Tourism Management, 2021
[paper] [journal version]
Press coverage: techFB techxplore

ViBE: Dressing for Diverse Body Shapes
Wei-Lin Hsiao, Kristen Grauman
CVPR, 2020 (Oral)
[paper][project page] [data] [CVPR oral talk]
Thanks CVPR Daily for covering our work!

Fashion++: Minimal Edits for Outfit Improvement
Wei-Lin Hsiao, Isay Katsman*, Chao-Yuan Wu*, Devi Parikh, Kristen Grauman
ICCV, 2019
[paper] [project page] [code]

Press coverage: voguebusiness wired venturebeat elle deeplearningAI fair engadget (and more)
Creating Capsule Wardrobes From Fashion Images
Wei-Lin Hsiao, Kristen Grauman
CVPR, 2018 (Spotlight Oral)
[paper] [data] [CVPR spotlight talk] [third-party blog 1] [third-party blog 2]

Learning the Latent "Look": Unsupervised Discovery of a Style-Coherent Embedding from Fashion Images
Wei-Lin Hsiao, Kristen Grauman
ICCV, 2017
[paper] [project page] [data] [code]


2020 - now     Visiting Researcher    
Facebook AI Research (FAIR), Austin, TX
Research on modeling the influence of culture on fashion.
2018 Research Intern
Facebook AI Research (FAIR), Menlo Park, CA
Working with Kristen Grauman and Devi Parikh.
Research on image generation for fashion (ICCV 2019).

Professional Activities

Organizer: UT Austin Vision Reading Group 2019-now. Contact me if you would like to join!
Member: Junior Graduate Admissions Committee 2018-2019
Participant: EECS Rising Stars, Berkeley, 2020
Reviewer: CVPR 2018, 2019, 2020, 2021
Reviewer: ECCV 2018, 2020
Reviewer: ICCV 2019
Teaching Assistance: Visual Recognition (CS 381V) 2017
Student Volunteer (Travel Grant): CVPR 2016

Press Coverage

CVPR Daily, Dressing for Diverse Body Shapes, 2020

Tech@facebook, Using AI to enhance tourism, 2020
TNW, Wake up sheeple: AI shows all tourists basically take the same picture, 2020
TechXplore, Facebook zeroes in on tourist photography habits, 2020
WWD, Inside Facebook’s Bid to Crack the Fashion AI Code, 2020
VentureBeat, Facebook details the AI behind its shopping experiences, 2020
UT-Austin News (Science & Tech), Artificial Intelligence System Gives Fashion Advice, 2019
WIRED, Skirt or Jeans? Now an AI Can Offer You Styling Tips, 2019
VentureBeat, Fashion++ turns your fashion Don’t into a Do with minimal tweaks, 2019
Vogue, Facebook experiments with AI-powered styling program, 2019
ELLE, Facebook Is Developing Artificial Intelligence That Offers Trend Advice To Make You More 'Fashionable', 2019
deepleaning.ai, Hope For the Fashion-Challenged, 2019
Facebook AI, Building AI to inform people's fashion choice, 2019
Engadget, Facebook’s latest AI experiment helps you pick what to wear, 2019


Computational Fashion Understanding

  • UT-Austin CS, June 2020
ViBE: Dressing for Diverse Body Shapes Creating Capsule Wardrobes From Fashion Images Learning the Latent "Look": Unsupervised Discovery of a Style-Coherent Embedding from Fashion Images

Last updated Feb. 2021.