AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Fine-Grained Segmentation Networks: Self-Supervised Segmentation for Improved Long-Term Visual Localization

Mans Larsson, Erik Stenborg, Carl Toft, Lars Hammarstrand, Torsten Sattler, Fredrik Kahl

ICCV 2019 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 74 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use FGSN with 100 clusters instead of 19 Cityscapes classesSSMC · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 71.8 -> 81.8Improved
use FGSN with 100 clusters instead of 19 Cityscapes classesSSMC · Extended CMU Seasons, Parklocalization % at 0.25 m / 2 deg: 32.8 -> 51.3Improved
use FGSN with 200 clusters instead of 19 Cityscapes classesSSMC · RobotCar Seasons, nightlocalization % at 0.25 m / 2 deg: 8.2 -> 9.6Improved
increase clusters from 100 to 1000 (ablation)SSMC with FGSN · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 81.8 -> 78.0Hurt
increase clusters from 100 to 1000 (ablation)SSMC with FGSN · Extended CMU Seasons, Parklocalization % at 0.25 m / 2 deg: 51.3 -> 45.1Hurt
initialize with segmentation pretraining instead of classification pretraining (200 clusters) (ablation)FGSN with SSMC · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 70.8 -> 81.0Improved
initialize with segmentation pretraining instead of classification pretraining (1000 clusters) (ablation)FGSN with SSMC · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 47.4 -> 78.0Improved
add correspondence loss (200 clusters) (ablation)FGSN with SSMC · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 78.7 -> 81.0Improved
add correspondence loss (1000 clusters) (ablation)FGSN with SSMC · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 73.4 -> 78.0Improved
add correspondence loss to 66-class Vistas segmentation network (ablation)SSMC with semantic classes · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 65.8 -> 66.5No clear effect
do not repeat/reset clustering during training (100 clusters) (ablation)FGSN with SSMC · Extended CMU Seasons, Urbanlocalization % at 0.25 m / 2 deg: 81.8 -> 85.3Improved
use FGSN 200 clusters in place of semantic segmentationPFSL · Extended CMU Seasons, Parklocalization % at 0.25 m / 2 deg: 39.0 -> 64.8Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy