AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Unsupervised Learning of Object Landmarks by Factorized Spatial Embeddings

James Thewlis, Hakan Bilen, Andrea Vedaldi

ICCV 2017 · Detection and segmentation › Open-set and few-shot detection · approach: Unsupervised object and part discovery · 145 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
increase number of unsupervised landmarks from 10 to 50 (ablation)Unsupervised landmarks · MAFL test (regressor on MAFL)mean error: 7.95 -> 6.67Improved
increase number of unsupervised landmarks from 10 to 50 (ablation)Unsupervised landmarks · MAFL test (regressor on CelebA)mean error: 6.32 -> 5.33Improved
train the regressor on CelebA instead of MAFL (ablation)Unsupervised landmarks, 50 landmarks · MAFL testmean error: 6.67 -> 5.33Improved
use all 19,000 MAFL training images instead of 10 (ablation)Unsupervised landmarks, 30 landmarks · MAFL testmean error: 8.49 -> 7.15Improved
use 20 labelled images instead of 1 (ablation)Unsupervised landmarks, 30 landmarks · MAFL testmean error: 10.82 -> 8.06Improved
add unsupervised pretraining before regressing on 5 AFLW images (ablation)Landmark regressor · AFLW (5 pts)mean error: 23.85 -> 12.94Improved
add unsupervised fine-tuning on target 300-W data (ablation)Unsupervised landmarks, 50 landmarks · 300-W (68 pts)mean error: 9.30 -> 7.97Improved
replace fully supervised detectors with unsupervised landmarks plus linear regressorUnsupervised landmarks, 50 points · MAFL testmean error: 7.95 -> 6.67Improved
replace fully supervised detectors with unsupervised landmarks plus linear regressorUnsupervised landmarks, 51 points · AFLW (5 pts)mean error: 7.23 -> 10.53Hurt
replace fully supervised RAR with unsupervised landmarks plus linear regressorUnsupervised landmarks, 50 landmarks · 300-W (68 pts)mean error: 4.94 -> 7.97Hurt

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