AI Paper Atlas

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

Miss Detection vs. False Alarm: Adversarial Learning for Small Object Segmentation in Infrared Images

Huan Wang, Luping Zhou, Lei Wang

ICCV 2019 · Detection and segmentation › Open-set and few-shot detection · approach: Open-set and unknown object detection · 534 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
use two adversarial generators for MD and FA instead of single CAN8 with combined objective (ablation)MDvsFA-cGAN · SingleF-measure: 0.31 -> 0.60Improved
use two adversarial generators instead of single UCAN64 with combined objective (ablation)MDvsFA-cGAN · SingleF-measure: 0.39 -> 0.60Improved
use two adversarial generators instead of single CAN8 with combined objective (ablation)MDvsFA-cGAN · AllSeqsF-measure: 0.20 -> 0.27Improved
use two adversarial generators instead of single UCAN64-cGAN (ablation)MDvsFA-cGAN · AllSeqsF-measure: 0.22 -> 0.27Improved
use two adversarial generators instead of single UCAN64-cGAN (ablation)MDvsFA-cGAN · SingleF-measure: 0.40 -> 0.60Improved
use different architectures (CAN8 and UCAN64) for the two generators instead of both CAN8 (ablation)MDvsFA-cGAN · AllSeqsF-measure: 0.17 -> 0.27Improved
use different architectures for the two generators instead of both CAN8 (ablation)MDvsFA-cGAN · SingleF-measure: 0.37 -> 0.60Improved
use different architectures for the two generators instead of both UCAN64 (ablation)MDvsFA-cGAN · AllSeqsF-measure: 0.20 -> 0.27Improved
use different architectures for the two generators instead of both UCAN64 (ablation)MDvsFA-cGAN · SingleF-measure: 0.50 -> 0.60Improved
vary lambda1 from 100 to 10 with lambda2=1 (ablation)MDvsFA-cGAN · SingleF-measure: 0.60 -> 0.58No clear effect
vary lambda2 from 10 down to 0.1 with lambda1=100 (ablation)MDvsFA-cGAN · SingleF-measure: 0.60 -> 0.60No clear effect
add adversarial MD/FA learning versus generic small-object segmentation FCN-MFBMDvsFA-cGAN · SingleF-measure: 0.47 -> 0.60Improved

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