AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

SemiContour: A Semi-Supervised Learning Approach for Contour Detection

Zizhao Zhang, Fuyong Xing, Xiaoshuang Shi, Lin Yang

CVPR 2016 · Label-efficient and robust learning › Semi-supervised learning · approach: Holistic semi-supervised pipelines · 48 citations (OpenAlex)

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
Incorporate unlabeled image patches to guide structured-random-forest node splittingSemiContour · BSDS500, 3 labeled imagesAP: SemiContour AP .776; SE-Var with 3 labeled images AP .74Improved
Use 3 labeled images rather than 200 (ablation)SemiContour · BSDS500ODS: .728 -> .736No clear effect
Use 10 labeled images rather than 3 (ablation)SemiContour · BSDS500ODS: .728 -> .732Improved
Use 20 labeled images rather than 3 (ablation)SemiContour · BSDS500ODS: .728 -> .734Improved
Use 50 labeled images rather than 3 (ablation)SemiContour · BSDS500ODS: .728 -> .736Improved
Use target-domain unlabeled data in cross-dataset trainingSemiContour · NYUD/BSDS cross-datasetOIS: .74 -> .75Improved
Use target-domain unlabeled data in cross-dataset trainingSemiContour · BSDS/NYUD cross-datasetODS: .64 -> .65Improved
Replace the original contour detector with SemiContour-Seg post-processingSemiContour · BSDS500ODS: ODS .73 -> .74; OIS .75 -> .77; AP .78 -> .76Mixed
Use fast sparse coding instead of iterative sparse-coding algorithmsSemiContour · Contour detection experimentscontour detection accuracyNo clear effect

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