AI Paper Atlas

Research map / Detection and segmentation / CNN features

Duality Diagram Similarity: a generic framework for initialization selection in task transfer learning

Kshitij Dwivedi, Jiahui Huang, Radoslaw Martin Cichy, Gemma Roig

ECCV 2020 · Detection and segmentation › CNN features · approach: Local features and context in CNNs · 25 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 z-score normalization (Q, D) to feature comparison (ablation)DDS (f=cosine) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.742 -> 0.864Improved
add z-score normalization (Q, D) to feature comparison (ablation)DDS (f=linear kernel) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.632 -> 0.842Improved
add z-score normalization (Q, D) to feature comparison (ablation)DDS (f=Laplacian) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.815 -> 0.860Improved
add z-score normalization (Q, D) to feature comparison (ablation)DDS (f=Pearson) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.823 -> 0.856Improved
use layer or group normalization instead of z-score (ablation)DDS (f=Laplacian) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.860 -> 0.806Hurt
add unbiased centering of similarity matrices (ablation)DDS (f=Laplacian) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.691 -> 0.860Improved
add unbiased centering of similarity matrices (ablation)DDS (f=euclidean) · Taskonomy winrate correlationSpearman correlation with transfer ranking: 0.613 -> 0.850Improved
use DDS (cosine) instead of RSA for model selectionTaskonomy model selection · Taskonomy affinity correlationSpearman correlation with transfer ranking: 0.777 -> 0.862Improved
use DDS instead of attribution-map methods (DeepLIFT)Taskonomy model selection · Taskonomy affinity correlationSpearman correlation with transfer ranking: 0.681 -> 0.862Improved
compute DDS from Pascal VOC images instead of Taskonomy images (ablation)DDS (f=Laplacian) · Pascal VOC segmentation transfer rankingSpearman correlation with transfer ranking: 0.5779 -> 0.765Improved
compute DDS from NYUv2 images instead of Pascal VOC images (ablation)DDS (f=Laplacian) · Pascal VOC segmentation transfer rankingSpearman correlation with transfer ranking: 0.765 -> 0.521Hurt
branch decoder from block 3 instead of block 1 (ablation)ResNet-50 encoder · NYUv2 depthlog RMSE (lower better): 0.322 -> 0.207Improved

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