AI Paper Atlas

Research map / Detection and segmentation / CNN features

Learning and Aggregating Deep Local Descriptors for Instance-level Recognition

Giorgos Tolias, Tomas Jenicek, Ondřej Chum

ECCV 2020 · Detection and segmentation › CNN features · approach: Local features and context in CNNs

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 fixed attention instead of learned attention (ablation)HOW ResNet18 · Validation (mini GLD2)mAP: 83.7 -> 84.8Improved
use contrastive metric-learning loss instead of cross entropy (ablation)HOW ResNet18 · Validation (mini GLD2)mAP: 75.5 -> 83.7Improved
use contrastive loss instead of cross entropy in full architecture (ablation)HOW ResNet18 · Revisited OxfordmAP: 69.9 -> 74.8Improved
use contrastive loss instead of cross entropy in full architecture (ablation)HOW ResNet18 · Tiny-GLD2muAP: 71.1 -> 81.3Improved
add local smoothing before ASMK aggregation (ablation)HOW ResNet18 · Validation (mini GLD2)number of descriptors per image: 344.1 -> 263.8Improved
use ResNet50 backbone instead of ResNet18HOW · Revisited Oxford/Paris (+R1M), mean over allmAP: 54.8 -> 58.0Improved
increase number of local features n from 1000 to 2000HOW ResNet50- · Revisited Oxford/Paris (+R1M), mean over allmAP: 58.0 -> 60.1Improved
use HOW local descriptors with ASMK instead of DELF with ASMKHOW ResNet50- · Revisited Oxford/Paris (+R1M), mean over allmAP: 53.0 -> 58.0Improved
use local descriptors with ASMK instead of best global descriptor GeM-APHOW ResNet50- · Revisited Oxford/Paris (+R1M), mean over allmAP: 49.9 -> 60.1Improved
use HOW ResNet50- instead of GEM_w on GLD2HOW · GLD2 classification (CLS3)muAP: 28.7 -> 36.5Improved

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