AI Paper Atlas

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

Visual Structure Constraint for Transductive Zero-Shot Learning in the Wild

Ziyu Wan, Dongdong Chen, Jing Liao

IJCV 2021 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 76 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 visual structure constraint with Wasserstein distanceVCL baseline · AwA2 (standard split, ResNet)MCA (%): 82.5 -> 96.7Improved
add visual structure constraint with Chamfer distanceVCL baseline · CUB (standard split, ResNet)MCA (%): 60.1 -> 74.2Improved
add structure constraint (Wasserstein) on proposed splitVCL baseline · AwA2 (proposed split)MCA (%): 61.5 -> 87.3Improved
add structure constraint (Wasserstein) in generalized ZSLVCL baseline · AwA2 (generalized ZSL)harmonic mean H (%): 34.6 -> 81.8Improved
add structure constraint (Wasserstein) in generalized ZSLVCL baseline · CUB (generalized ZSL)harmonic mean H (%): 26.5 -> 57.5Improved
replace hard one-to-one bipartite matching with soft Wasserstein couplingBMVSc · AwA2 (proposed split)MCA (%): 81.7 -> 87.3Improved
replace Chamfer distance with bipartite matchingCDVSc · CUB (generalized ZSL)harmonic mean H (%): 51.4 -> 47.9Hurt
apply structure constraint directly on test set mixed with unrelated images (ablation)WDVSc · AwA2 (standard split, with 8K unrelated images)MCA (%): 82.5 -> 81.3Hurt
add filtering of unrelated images (S+) before the constraint (ablation)WDVSc · AwA2 (standard split, with 8K unrelated images)MCA (%): 81.3 -> 92.4Improved
add filtering of unrelated images (S+) before the constraint (ablation)WDVSc · AwA2 (proposed split, with 8K unrelated images)MCA (%): 60.8 -> 78.3Improved
increase number of K-means clusters from 3 to 10 (ablation)WDVSc · AwA2 (proposed split)MCA (%): 63.4 -> 87.3Improved
increase number of K-means clusters from 3 to 10 (ablation)BMVSc · AwA2 (proposed split)MCA (%): 62.3 -> 81.7Improved

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