Georgy Ponimatkin

I am a Ph.D. student at the Inteligent Machine Perception lab at CIIRC CTU, supervised by Josef Sivic.

I obtained my master's degree in high energy physics at CTU Prague, where I worked at the STAR experiment at the Brookhaven National Lab. I also spent two and half years at the IMAGINE team at Ecole des Ponts ParisTech.

Email  /  CV  /  Google Scholar  /  Github

Georgy Ponimatkin
Research

Currently my research is focused on Vision-Language-Action models and ways to improve their generalizability for robotic manipulation. Previously I worked mainly on 6D pose estimation and unsupervised video object segmentation.

6D Object Pose Tracking in Internet Videos for Robotic Manipulation
Georgy Ponimatkin*, Мartin Cífka*, Tomáš Souček, Médéric Fourmy, Yann Labbé, Vladimír Petrík, Josef Sivic
ICLR 2025
paper / arxiv / code

A method to estimate 6D pose of the object in the wild given an approximate mesh.
FocalPose++: Focal Length and Object Pose Estimation via Render and Compare
Мartin Cífka* Georgy Ponimatkin*, Yann Labbé, Bryan Russell, Mathieu Aubry, Vladimír Petrík, Josef Sivic
TPAMI 2024
paper / arxiv / code

An extended render and compare method for 6D pose estimation in uncalibrated settings.
A Simple and Powerful Global Optimization for Unsupervised Video Object Segmentation
Georgy Ponimatkin, Nermin Samet, Yang Xiao, Yuming Du, Renaud Marlet, Vincent Lepetit
WACV 2023
paper / arxiv / code

Spectral clustering method for unsupervised video object segmentation.
Focal Length and Object Pose Estimation via Render and Compare
Georgy Ponimatkin, Yann Labbé, Bryan Russell, Mathieu Aubry, Josef Sivic
CVPR 2022
paper / arxiv / code

A render and compare method for 6D pose estimation in uncalibrated settings.
NOPE: Novel Object Pose Estimation from a Single Image
Van Nguyen Nguyen, Thibault Groueix, Georgy Ponimatkin, Yinlin Hu, Renaud Marlet, Mathieu Salzmann, Vincent Lepetit,
CVPR 2024
paper / arxiv / code

A single reference image relative pose estimator for unseen objects.
You Never Get a Second Chance To Make a Good First Impression: Seeding Active Learning for 3D Semantic Segmentation
Nermin Samet, Oriane Simeoni, Gilles Puy, Georgy Ponimatkin, Renaud Marlet, Vincent Lepetit,
ICCV 2023
paper / arxiv / code

An active learning method for point cloud semantic segmentation.
CNOS: A Strong Baseline for CAD-based Novel Object Segmentation
Van Nguyen Nguyen, Thibault Groueix, Georgy Ponimatkin, Vincent Lepetit, Tomas Hodan
ICCV R6D Workshop 2023
paper / arxiv / code

An unseen object detector baseline for BOP Challenge.
Reviewing
Reviewing CVPR, ECCV, WACV, ACCV, 3DV, NeurIPS, ICLR

This website is based on the template provied by Jon Barron.