AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Dynamic Conditional Networks for Few-Shot Learning

Fang Zhao, Jian Zhao, Shuicheng Yan, Jiashi Feng

ECCV 2018 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Fine-grained and cross-domain few-shot baselines · 75 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
replace plain FC conditional layer with 8-D dynamic conditional layerResNet-34 conditional net · Amazon Bin Image identificationaccuracy: 74.48 -> 76.81Improved
replace plain FC conditional layer with 8-D dynamic layerResNet-34 conditional net · Amazon Bin Image verificationaccuracy: 84.87 -> 85.48Improved
increase dynamic layer dimension from 8 to 16 (ablation)DCCN · Amazon Bin Image identificationaccuracy: 76.81 -> 75.66Hurt
increase dynamic layer dimension from 4 to 8 (ablation)DCCN · Amazon Bin Image identificationaccuracy: 76.60 -> 76.81Improved
replace plain network with 64-D dynamic layerResNet-34 multi-modal classifier · MIRFlickr-25K 10k tagsIoU: 0.6516 -> 0.6606Improved
replace plain network with 64-D dynamic layerResNet-34 multi-modal classifier · MIRFlickr-25K 5k tagsIoU: 0.6489 -> 0.6553Improved
increase dynamic layer dimension from 64 to 128 (ablation)DCCN · MIRFlickr-25K 10k tagsIoU: 0.6606 -> 0.6516Hurt
use 16-D dynamic layer instead of fully connected layersGroundeR-style grounding · Flickr30k Entitiesaccuracy: 47.81 -> 50.65Improved
increase dynamic layer dimension from 16 to 32 (ablation)DCCN · Flickr30k Entitiesaccuracy: 50.65 -> 50.52No clear effect
dynamic layer versus GroundeR on People phrasesGroundeR · Flickr30k Entities Peopleaccuracy: 61.00 -> 67.37Improved
dynamic layer versus GroundeR on Vehicles and Scene phrasesGroundeR · Flickr30k Entities Vehicles/Sceneaccuracy: 68.75 -> 56.04; 58.18 -> 54.05Hurt
use dynamic conditional layers in conditional GANplain conditional GAN · MS-Celeb-1M face generationrank-1 accuracy: 0.457 -> 0.688Improved

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