AI Paper Atlas

Research map / Detection and segmentation / CNN features

ZMIS-SAM: Segment Anything Model Enhanced With Wavelet Transform For Zooplankton Microscopy Image Instance Segmentation

Dekun Yuan, Zhongwei Li, Zheng Qiao, Jie Zhang

ECCV 2026 · Detection and segmentation › CNN features · approach: Domain-specific segmentation and detection · 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
replace ViT with ZM-ViT (shape and intensity adapters) (ablation)USIS-SAM ViT-H · ZMIS5KmAP: 70.4 -> 71.7Improved
add NFAM on top of ZM-ViT (ablation)ZMIS-SAM · ZMIS5KmAP: 71.7 -> 73.0Improved
add WM2FE on top of ZM-ViT (ablation)ZMIS-SAM · ZMIS5KmAP: 71.7 -> 72.7Improved
add both NFAM and WM2FE (ablation)ZMIS-SAM · ZMIS5KAP75: 77.7 -> 80.7Improved
use Haar wavelet instead of rbio1.1 or bior1.1 (ablation)WM2FE · ZMIS5KmAP: 72.0 / 72.7 -> 73.6Improved
add Shape Adapter alone (ablation)ZM-ViT · ZMIS5KmAP: 70.4 -> 71.2Improved
add Intensity Adapter alone (ablation)ZM-ViT · ZMIS5KmAP: 70.4 -> 71.0Improved
use 4 MConvBlocks rather than 1 (ablation)NFAM · ZMIS5KmAP: 72.2 -> 73.0Improved
use 6 MConvBlocks rather than 4 (ablation)NFAM · ZMIS5KmAP: 73.0 -> 72.7No clear effect
add WM2FE to USIS-SAM on another datasetUSIS-SAM · USIS10KmAP: 34.6 -> 35.4Mixed
ZMIS-SAM vs RSPrompter inferenceZMIS-SAM · inference speedFPS: 8.8 -> 3.9Hurt
ZMIS-SAM vs best traditional model Mask2FormerZMIS-SAM · ZMIS5KAP50: 87.4 -> 94.6Improved

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