AI Paper Atlas

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

SHERPA: Fine-tuning Segment Anything Models with Task-relevant Guidance

Jingcheng Xie, Yinda Chen, Xiaoyu Liu, Haoyuan Shi, Zhiwei Xiong

ICML 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Interactive and referring segmentation

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 FRS module to small-model guidance (ablation)SHERPA · DISK-5KF1 general: 0.5752 -> 0.7930Improved
add FRS module to small-model guidance (ablation)SHERPA · DUTSF1 general: 0.6432 -> 0.8284Improved
add FRS module to small-model guidance (ablation)SHERPA · LucchiF1 general: 0.7611 -> 0.7857Improved
add FRS module to small-model guidance (ablation)SHERPA · DISK-5KF1 valid: 0.8695 -> 0.8855Improved
use small SAM instead of large SAM itself as guide (ablation)SHERPA · DISK-5KF1 general: 0.7797 -> 0.7930Improved
fine-tune small guiding model with KL constraint (ablation)SHERPA · DISK-5KF1 general: 0.7812 -> 0.7930Improved
use task-relevant feature guidance instead of regularizing residual components (ablation)SHERPA · DISK-5KF1 general: 0.7431 -> 0.7930Improved
set Var Ratio to 0.75 instead of 1.00 (all channels) (ablation)SHERPA · DISK-5KF1 general: 0.5257 -> 0.7930Improved
set Var Ratio to 0.75 instead of 0.00 (ablation)SHERPA · DISK-5KF1 general: 0.7234 -> 0.7930Improved
use SHERPA instead of standard fine-tuningSAM ViT-H · DISK-5k generalization retentionretention: 85.3% -> 93.4%Improved
use SHERPA instead of standard fine-tuningSAM ViT-H · Lucchi generalization retentionretention: 78.9% -> 91.8%Improved
use SHERPA instead of standard fine-tuningSAM2 · VOST generalization retentionretention: 84.0% -> 95.1%Improved

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