Mingqi Gao
Scholar

Mingqi Gao

Google Scholar ID: ECCd0hwAAAAJ
Research Associate, University of Sheffield
Video SegmentationMulti-Modal Large Models3D Vision
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
905
 
H-index
13
 
i10-index
14
 
Publications
20
 
Co-authors
0
 
Contact
Emailim.mingqi@gmail.comGitHubOpen ↗LinkedInOpen ↗
Publications
6 items
CIRSeg: Coarse-to-Fine Intensity-Robust Liver Segmentation with Source-Free Continual Test-Time Adaptation
2026
Cited
0
Competitive Memory Readout for Robust Video Object Segmentation: 2nd Place Technical Report for the MOSEv2 Track of the 8th LSVOS Challenge
2026
Cited
0
DeSeG: Decoupling Semantic Intent and Geometric Constraints for Physically Plausible Human-Scene Interaction
2026
Cited
0
Reinforcing 3D Understanding in Point-VLMs via Geometric Reward Credit Assignment
2026
Cited
0
UniSurgSAM: A Unified Promptable Model for Reliable Surgical Video Segmentation
2026
Cited
0
Re-Prompting SAM 3 via Object Retrieval: 3rd of the 5th PVUW MOSE Track
2026
Cited
0
Resume
Academic Achievements
  • 1st Place, EPIC-KITCHENS Challenge (Semi-Supervised VOS Track), CVPR 2025
  • 1st Place, LSVOS Challenge on Complex Video Object Segmentation (MOSEv2), ICCV 2025
  • 1st Place, PVUW Challenge (MeViS Track), CVPR 2024
  • 1 paper on Referring Video Object Segmentation accepted by ICCV 2023
  • 2 papers on Referring and Visible-Thermal Video Object Segmentation published in IEEE TCSVT 2023
  • Survey paper on video object segmentation published in Artificial Intelligence Review, 2022
  • Paper on point-based memory matching network for VOS published in Pattern Recognition, 2022
  • Work on texture segmentation using feature fusion and NMF published in Signal Processing, 2019
  • Factorization-based active contour model for texture segmentation presented at ICIP 2016
  • Released open-source project Track-Anything (6869+ stars on GitHub)