Research & Work Experience

Intel Labs

Staff Research Scientist AI/ML, 2021-present
Santa Clara, CA USA

• Currently working on video understanding, especially egocentric video understanding

• Leading multiple research projects on efficient video representations using sparse transformers and graph neural networks

• Actively collaborating with several academic partners on diffusion models and machine unlearning

• Led the project of GraVi-T, which is an advanced graph learning framework for efficient video understanding

• Won several challenges on video understanding algorithms, including Ego4D and ActivityNet

University of Michigan
Graduate Student Research Assistant, 2017-2021
Ann Arbor, MI USA

• Conducted research on video understanding and spatiotemporal action detection

• Developed 3D object detection algorithms using deep learning that use a sequence of images and LiDAR points for the group project in course EECS 599: Self-driving cars

• Performed research on automatic image colorization with GAN for the group project in course EECS 545: Machine learning

• Developed standalone executable programs for evaluating various video segmentation algorithms for DARPA's Media Forensics (MediFor) project teams

Seoul National University
Summer Intern, 2015-2016
Seoul South Korea

• Performed research on VQA and implemented multiple deep learning algorithms for high-performing VQA systems

Xeron
Video Analytics Algorithm Developer, 2012-2015
Seoul South Korea

• Developed video analytics algorithms for the monitoring system of the hospital's neonatal units that detect potentially dangerous situations

• Developed multiple video analytics algorithms and built a state-of-the-art event detection system for the i-LIDS benchmark, which was certified by UK Home Office

Publications

• R.A.C.E.: Robust Adversarial Concept Erasure for Secure Text-to-Image Diffusion Model
  European Conference on Computer Vision (ECCV), 2024. Oral Presentation
  Changhoon Kim*, Kyle Min*, Yezhou Yang  (*:  equal contribution)
  [ paper ]

• WOUAF: Weight Modulation for User Attribution and Fingerprinting in Text-to-Image Diffusion Models
  IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2024.
  Preliminary version was presented at the NeurIPS Workshop on Diffusion Models (NeurIPSW), 2023.
  Changhoon Kim*, Kyle Min*, Maitreya Patel, Sheng Cheng, Yezhou Yang  (*:  equal contribution)
  [ paper | code ]

• Action Scene Graphs for Long-Form Understanding of Egocentric Videos
  IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2024.
  Ivan Rodin*, Antonino Furnari*, Kyle Min*, Subarna Tripathi, Giovanni Maria Farinella  (*:  equal contribution)
  [ paper | code ]

• STHG: Spatial-Temporal Heterogeneous Graph Learning for Advanced Audio-Visual Diarization
  IEEE Conference on Computer Vision and Pattern Recognition Workshop on Ego4D (CVPRW), 2023. Spotlight Presentation
  Kyle Min
  [ tech report | code ]

• SViTT: Temporal Learning of Sparse Video-Text Transformers
  IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2023.
  Yi Li, Kyle Min, Subarna Tripathi, Nuno Vasconcelos
  [ paper | code ]

• Unbiased Scene Graph Generation in Videos
  IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2023.
  Sayak Nag, Kyle Min, Subarna Tripathi, Amit K Roy-Chowdhury
  [ paper | code ]

• Intel Labs at Ego4D Challenge 2022: A Better Baseline for Audio-Visual Diarization
  European Conference on Computer Vision Workshop on Ego4D (ECCVW), 2022.
  Kyle Min
  [ tech report | code ]

• Learning Long-Term Spatial-Temporal Graphs for Active Speaker Detection
  European Conference on Computer Vision (ECCV), 2022.
  Kyle Min*, Sourya Roy*, Subarna Tripathi, Tanaya Guha, Somdeb Majumdar  (*:  equal contribution)
  [ paper | code | poster | presentation ]

• Intel Labs at ActivityNet Challenge 2022: SPELL for Long-Term Active Speaker Detection
  IEEE Conference on Computer Vision and Pattern Recognition Workshop on Ego4D (CVPRW), 2022.
  Kyle Min, Sourya Roy, Subarna Tripathi, Tanaya Guha, Somdeb Majumdar
  [ tech report | presentation ]

• Integrating Human Gaze into Attention for Egocentric Activity Recognition
  IEEE Winter Conference on Applications of Computer Vision (WACV), 2021.
  Kyle Min, Jason J. Corso
  [ paper | code | presentation ]

• Adversarial Background-Aware Loss for Weakly-supervised Temporal Activity Localization
  European Conference on Computer Vision (ECCV), 2020.
  Kyle Min, Jason J. Corso
  [ paper | code | poster | presentation ]

• TASED-Net: Temporally-Aggregating Spatial Encoder-Decoder Network for Video Saliency Detection
  IEEE International Conference on Computer Vision (ICCV), 2019.
  Kyle Min, Jason J. Corso
  [ paper | code ]

• Hierarchical Novelty Detection for Visual Object Recognition
  IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018.
  Kibok Lee, Kimin Lee, Kyle Min, Yuting Zhang, Jinwoo Shin, and Honglak Lee
  [ paper | code ]

Awards and Honors

• 1st place in the Ego4D Challenge for Audio-Visual Diarization ($3000)
  Ego4D sponsored by Meta, ECCVW 2022

• 2nd place in the AVA Challenge for Active Speaker Detection
  ActivityNet, CVPRW 2022

• Best Achievement Award of Global Top Talent Forum ($3000)
  Hyundai Motor Group, 2019

• Richard F. and Eleanor A. Towner Prize for Distinguished Academic Achievement ($1000)
  University of Michigan, 2018

• College of Engineering Alumni Association Prize for the best academic achievement ($1000)
  Seoul National University, 2016

• B.S. summa cum laude
  Seoul National University, 2016

• Stipends for outstanding students ($8000)
  Korea Foundation for Advanced Studies, 2011-2012

• National Science & Technology scholarship (Full tuition)
  Korea Student Aid Foundation, 2009-2012

• Scholarship for Distinguished Academic Achievement ($1000)
  Samsung Electronics, 2008

Teaching Experience

• Graduate Student Instructor, EECS 442 (Computer Vision)
  University of Michigan, Fall 2017

• Freshman adviser, Physics and Mathematics
  Seoul National University, Fall 2011

Professional Activities

• Reviewer for CVPR / ECCV / ICLR / ICML / NeurIPS, 2024

• Reviewer for two manuscripts of IEEE TIP, 2023-2024

• Reviewer for two manuscripts of IEEE TMM, 2023-2024

• Reviewer for CVPR / ICCV / ICLR / ICML / NeurIPS, 2023

• Reviewer for CVPR / ECCV / ICLR / NeurIPS, 2022

• Reviewer for manuscripts of IEEE TPAMI / Pattern Recognition, 2021

• Reviewer for CVPR / ICCV, 2021

• Reviewer for ECCV, 2020

• Invited talk for the SCAI AI Day at Arizona State University (Phoenix, AZ), 2023

• Spotlight presentation at Ego4D workshop in CVPR (Vancouver), 2023

• Hyundai Global Top Talent Forum - (San Diego, CA), 2019

Other Activities

• Poster presentation at Intel Innovation (San Jose, CA), 2022

• Poster presentation at EECS Homecoming Luncheon, 2019

• Student volunteer for ICCV (Seoul, Korea), 2019

• Video demo at ECE Open House, 2019

• Qualcomm IT Tour (San Diego, CA), 2015

• LG Global Challenger (England, Sweden, Belgium, Netherlands), 2011