AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

PINet: Improving the Stability of Prototype Networks via Phantasia-Inspired Uncertain Representations

Ho Kyung Shin, Soeun Bae, Sang Min Kim, Byoung Chul Ko, Woo-Jeoung Nam

AAAI 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Prototype and concept dictionary models · 0 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 coarse location guidance to the empty-input feature extraction (ablation)PINet (ResNet-152) · CUB-200-2011accuracy (%): 50.4 -> 85.3Improved
add coarse location guidance (ablation)PINet (ResNet-152) · CUB-200-2011consistency score: 36.6 -> 62.3Improved
use certain prototype instead of uncertain prototype at inference (ablation)PINet (DenseNet-161) · CUB-200-2011accuracy (%): 83.1 -> 86.7Improved
apply PINet to ViT-S instead of CNN backbonesPINet · CUB-200-2011consistency score: ViT-S 46.8 vs ResNet-152 62.3Hurt
add PINet uncertain representationsProtoPNet-style models (ResNet-34) · CUB-200-2011 noise sigma=0.2accuracy (%): 74.9 -> 80.4Improved
add PINet uncertain representationsProtoPNet-style models (ResNet-152) · CUB-200-2011 noise sigma=0.2consistency score: 32.3 -> 57.4Improved

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