arXiv:2606.12869v1 Announce Type: new 
Abstract: In image and surface-based learning tasks, convolutional features are typically extracted using receptive fields that are sampled uniformly across the entire domain. However, informative structures are rarely distributed uniformly in practice and are often concentrated in localized regions. Such phenomena are particularly common in medical imaging, where pathological changes are spatially confined. Consequently, uniform convolution allocates equal computational effort to both informative and uninformative regions, resulting in inefficient feature extraction and suboptimal utilization of model capacity. To address this issue, we propose a framework for task-adaptive sampling that dynamically redistributes computational attention according to the spatial importance of the data. Specifically, we introduce the Density-Equalizing Convolutional Neural Network (DECNN), which employs density-equalizing mappings to guide convolution through a learned density function. The density function encodes the relative importance of different regions and induces a transformation that enlarges informative areas while compressing less relevant ones. As a result, convolutional receptive fields are redistributed non-uniformly over the domain, enabling denser sampling in task-relevant regions. By coupling this importance-driven transformation with convolution, DECNN performs adaptive feature extraction that focuses computational resources on informative structures. This leads to more efficient use of model capacity, yielding a lightweight yet expressive architecture while simultaneously producing an interpretable saliency map. Experiments on image classification and craniofacial surface analysis demonstrate that DECNN achieves competitive or superior performance with fewer parameters, accurately identifies task-relevant regions, and remains robust under complex geometric variations.

تم اقتراح إطار جديد لأخذ العينات المتكيف مع المهام في التصوير الطبي، مع التركيز على الشبكة العصبية التلافيفية المعادلة للكثافة (DECNN). تهدف هذه الطريقة إلى تحسين استخراج الميزات من خلال إعادة توزيع الانتباه الحسابي بناءً على الأهمية المكانية للبيانات، مما يعالج عدم الكفاءة في طرق أخذ العينات الموحدة التقليدية.

Se ha propuesto un nuevo marco para el muestreo adaptativo a tareas en la imagen médica, centrándose en la Red Neuronal Convolucional de Igualación de Densidad (DECNN). Este enfoque busca optimizar la extracción de características redistribuyendo la atención computacional según la importancia espacial de los datos, abordando ineficiencias en los métodos de muestreo uniforme tradicionales.

Un nouveau cadre pour l'échantillonnage adaptatif aux tâches dans l'imagerie médicale a été proposé, mettant l'accent sur le Réseau de Neurones Convolutionnels Égalisant la Densité (DECNN). Cette approche vise à optimiser l'extraction des caractéristiques en redistribuant l'attention computationnelle en fonction de l'importance spatiale des données, abordant ainsi les inefficacités des méthodes d'échantillonnage uniformes traditionnelles.

A new framework for task-adaptive sampling in medical imaging has been proposed, focusing on the Density-Equalizing Convolutional Neural Network (DECNN). This approach aims to optimize feature extraction by redistributing computational attention based on the spatial importance of data, addressing inefficiencies in traditional uniform sampling methods.