AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Feature Weighting and Boosting for Few-Shot Segmentation

Khoi Nguyen, Sinisa Todorovic

ICCV 2019 · Detection and segmentation › Open-set and few-shot detection · approach: Few-shot segmentation · 411 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 feature weighting (contribution 1) (ablation)B, VGG 16 · PASCAL-5i one-shotmIoU: 47.55 -> 48.68Improved
add feature weighting (contribution 1) (ablation)B, ResNet 101 · PASCAL-5i one-shotmIoU: 51.30 -> 52.71Improved
add feature boosting with an ensemble of experts (contribution 2) (ablation)B, VGG 16 · PASCAL-5i one-shotmIoU: 47.55 -> 51.22Improved
add feature boosting with an ensemble of experts (contribution 2) (ablation)B, ResNet 101 · PASCAL-5i one-shotmIoU: 51.30 -> 54.21Improved
add feature weighting (contribution 1) (ablation)B+C2, VGG 16 · PASCAL-5i one-shotmIoU: 51.22 -> 51.90Improved
add feature weighting (contribution 1) (ablation)B+C2, ResNet 101 · PASCAL-5i one-shotmIoU: 54.21 -> 56.19Improved
add feature weighting (contribution 1) (ablation)B, VGG 16 · COCO-20i one-shotmIoU: 13.26 -> 17.11Improved
add feature weighting (contribution 1) (ablation)B, ResNet 101 · COCO-20i one-shotmIoU: 17.23 -> 18.86Improved
add feature boosting with an ensemble of experts (contribution 2) (ablation)B, VGG 16 · COCO-20i one-shotmIoU: 13.26 -> 17.73Improved
add feature boosting with an ensemble of experts (contribution 2) (ablation)B, ResNet 101 · COCO-20i one-shotmIoU: 17.23 -> 19.56Improved
jointly analyze K support images instead of averaging per-support predictions (ablation)B+C1+C2, VGG 16 · PASCAL-5i five-shotmIoU: 52.62 -> 55.08Improved
jointly analyze K support images instead of averaging per-support predictions (ablation)B+C1+C2, ResNet 101 · PASCAL-5i five-shotmIoU: 57.78 -> 59.92Improved

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