Model Agnostic Preference Optimization for Medical Image Segmentation
By: Yunseong Nam , Jiwon Jang , Dongkyu Won and more
Potential Business Impact:
Teaches computers to see body parts better.
Preference optimization offers a scalable supervision paradigm based on relative preference signals, yet prior attempts in medical image segmentation remain model-specific and rely on low-diversity prediction sampling. In this paper, we propose MAPO (Model-Agnostic Preference Optimization), a training framework that utilizes Dropout-driven stochastic segmentation hypotheses to construct preference-consistent gradients without direct ground-truth supervision. MAPO is fully architecture- and dimensionality-agnostic, supporting 2D/3D CNN and Transformer-based segmentation pipelines. Comprehensive evaluations across diverse medical datasets reveal that MAPO consistently enhances boundary adherence, reduces overfitting, and yields more stable optimization dynamics compared to conventional supervised training.
Similar Papers
SAMPO: Visual Preference Optimization for Intent-Aware Segmentation with Vision Foundation Models
CV and Pattern Recognition
Helps computers find all similar things, not just one.
VAGPO: Vision-augmented Asymmetric Group Preference Optimization for the Routing Problems
Machine Learning (CS)
Solves huge delivery routes fast without retraining
Self-NPO: Negative Preference Optimization of Diffusion Models by Simply Learning from Itself without Explicit Preference Annotations
CV and Pattern Recognition
Teaches AI to avoid making bad pictures.