Building scalable 3D understanding with minimal supervision.

Computer Vision PhD Student at CTU Prague.

Jan Skvrna

Winner of S23DR Challenge 2025

Secured 1st place (and $10k prize) at the CVPR 2025 Structured Semantic 3D Reconstruction Challenge.

View Leaderboard

About Me

I am a PhD student at the Visual Recognition Group, Czech Technical University, advised by Lukas Neumann.

My research focuses on Weakly-Supervised 3D Detection for autonomous driving. I aim to reduce the reliance on expensive 3D annotations by leveraging temporal consistency and 2D cues from off-the-shelf vision models.

Previously, I completed my MSc with honours, receiving the Dean's Award for my thesis on 3D object detection.

2 Core A* Papers
1st CVPR Challenge

Selected Publications

MonoSOWA Poster
ICCV 2025

MonoSOWA: Scalable monocular 3D Object detector Without human Annotations

A novel approach enabling 3D object detection from a single RGB camera without human annotations. Introduces a Local Object Motion Model with ~700× speedup over prior work.

TCC-Det Poster
ECCV 2024

TCC-Det: Temporarily consistent cues for weakly-supervised 3D detection

Leveraging temporal consistency to train 3D detectors without 3D labels. Outperforms prior weakly-supervised methods on KITTI and Waymo datasets.