AI Paper Atlas

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

Task-Driven Modular Networks for Zero-Shot Compositional Learning

Senthil Purushwalkam, Maximilian Nickel, Abhinav Gupta, Marc'Aurelio Ranzato

ICCV 2019 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 174 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 task-driven gatings (ablation)TMN · MIT-StatesTop-1 val AUC x100: 3.2 -> 3.5Improved
add task-driven gatings (ablation)TMN · UT-ZapposTop-1 val AUC x100: 32.7 -> 36.8Improved
add joint extraction of image and pair features (embedding at input rather than output) (ablation)TMN without task-driven gatings · MIT-StatesTop-1 val AUC x100: 0.8 -> 3.2Improved
add joint extraction of image and pair features (embedding at input rather than output) (ablation)TMN without task-driven gatings · UT-ZapposTop-1 val AUC x100: 20.1 -> 32.7Improved
add ConceptDrop regularization (ablation)TMN · MIT-StatesTop-1 val AUC x100: 3.3 -> 3.5Improved
add ConceptDrop regularization (ablation)TMN · UT-ZapposTop-1 val AUC x100: 35.7 -> 36.8Improved
increase layers from 1 to 3 (24 modules) (ablation)TMN · MIT-States valVal AUC x100: 2.50 -> 3.51Improved
increase layers from 3 to 5 (24 modules) (ablation)TMN · MIT-States valVal AUC x100: 3.51 -> 3.24Hurt
increase modules per layer from 12 to 24 (3 layers) (ablation)TMN · MIT-States valVal AUC x100: 3.23 -> 3.51Improved
increase modules per layer from 12 to 30 (5 layers) (ablation)TMN · MIT-States valVal AUC x100: 3.48 -> 3.19Hurt
use TMN instead of RedWineRedWine · MIT-States testTop-1 test AUC x100: 2.4 -> 2.9Improved
use TMN instead of RedWineRedWine · UT-Zappos testTop-1 test AUC x100: 27.1 -> 29.3Improved

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