AI Paper Atlas

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

On Equivariant and Invariant Learning of Object Landmark Representations

Zezhou Cheng, Jong-Chyi Su, Subhransu Maji

ICCV 2021 · Detection and segmentation › Open-set and few-shot detection · approach: Object-centric learning theory and video · 15 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
add linear projection trained with spatial contrastive loss to hypercolumn (ablation)Ours hypercolumn · MAFL different-identity matching (aligned)mean pixel error: 6.16 -> 2.06Improved
add linear projection to hypercolumn (ablation)Ours hypercolumn · MAFL different-identity matching (in-the-wild)mean pixel error: 5.58 -> 3.03Improved
add linear projection to hypercolumn (ablation)Ours hypercolumn · MAFL same-identity matching (aligned)mean pixel error: 0.73 -> 0.71No clear effect
stack hypercolumn blocks #2-#5 instead of single block #4 (ablation)Ours ResNet50 · AFLW_M landmark detectioninter-ocular distance error: 8.83 -> 6.99Improved
stack hypercolumn blocks #2-#5 instead of single block #4 (ablation)Ours ResNet50 · MAFL landmark detectioninter-ocular distance error: 2.73 -> 2.44Improved
use deeper block 4 instead of block 1 features (ablation)Ours ResNet50 · AFLW_M landmark detectioninter-ocular distance error: 24.20 -> 8.83Improved
add contrastive pretraining instead of random initialization (ablation)ResNet50 hypercolumn · AFLW_M landmark detectioninter-ocular distance error: 16.74 -> 6.99Improved
add contrastive pretraining instead of ImageNet pretraining (ablation)ResNet50 hypercolumn · AFLW_M landmark detectioninter-ocular distance error: 8.88 -> 6.99Improved
fine-tune whole network (ablation)Ours block 4 ResNet50 · MAFL / AFLW_M landmark detectioninter-ocular distance error: 2.73 -> 2.81 (MAFL), 8.83 -> 7.80 (AFLW_M)Mixed
replace NMF with learned projection for 64-D reduction (ablation)Ours · AFLW_M landmark detectioninter-ocular distance error: 7.60 -> 7.21Improved
use ResNet50 instead of ResNet18 (ablation)Ours hypercolumn · AFLW_M landmark detectioninter-ocular distance error: 8.59 -> 6.99Improved
train on 128x128 images instead of 96x96 (ablation)Ours ResNet50 · CUB landmark detectionPCK: 68.63 -> 72.61Improved

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