arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新，休息一下吧，周末愉快~~
arXiv 2603.17879cs.CVcs.AI

解剖引导的视觉-语言学习与角度原型分离用于类别不平衡下的多标签视频胶囊内镜分类

Anatomy-Guided Vision-Language Learning with Angular Prototype Separation for Multi-Label Video Capsule Endoscopy Classification Under Class Imbalance

Department of Engineering Science, IIT Hyderabad（印度海得拉尔理工学院工程科学系）；Department of Biomedical Engineering, IIT Hyderabad（印度海得拉尔理工学院生物医学工程系）

Podakanti Satyajith Chary, Nagarajan Ganapathy

备注：
12 pages, 1 figure, ICPR 2026 RARE-VISION Competition
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对视频胶囊内镜多标签分类中的极端类别不平衡问题，提出结合角度分离损失和生物状态机解码器的框架，利用解剖上下文和原型学习提升罕见类检测性能。

摘要：

本文提出一个多标签时间事件检测框架用于视频胶囊内镜（VCE），通过结合两个主要贡献来解决Galar数据集固有的极端类别不平衡问题：类原型上的角度分离损失和生物状态机时间解码器。主干网络保持为BiomedCLIP，一个生物医学视觉-语言基础模型。三个连续帧通过局部差分注意力模块融合，该模块通过抑制静态时间冗余来放大瞬态病理信号。然后，解剖上下文头将病理预测条件化于软解剖激活上，利用已知的胃肠道发现空间共现结构。可学习的文本特征提示和基于原型的logit增强与角度分离损失一起训练，该损失惩罚类原型之间的非对角线余弦相似度，防止在极端不平衡下影响罕见类的原型崩溃。为抵消倾斜的标签分布，训练方案结合了非对称焦点损失、逆频率加权采样、时间混合、指数移动平均和每类阈值校准。生物状态机解码器用基于解剖标签的生理学基础前向状态转换替代朴素间隙合并，消除了先前方法中每视频产生数百个虚假解剖事件的碎片化伪影，并将每视频解剖输出减少到2-3个临床现实事件。在包含三个NaviCam检查（161,025帧）的保留RARE-VISION测试集上，更新后的管道实现了整体时间mAP@0.5为0.3597，mAP@0.95为0.3399，相比先前提交分别相对提升46%和44%，总推理时间在单个GPU上约21分钟完成。

英文摘要：

This work presents a multi-label temporal event detection framework for video capsule endoscopy (VCE) that addresses the extreme class imbalance inherent in the Galar dataset by combining two principal contributions: an Angular Separation Loss on class prototypes and a Biological State Machine temporal decoder. The backbone remains BiomedCLIP, a biomedical vision-language foundation model. Three consecutive frames are fused through a Local Differencing Attention module that amplifies transient pathological signals by suppressing static temporal redundancy. An Anatomy Context Head then conditions pathological predictions on soft anatomical activations, exploiting the known spatial co-occurrence structure of GI findings. Learnable text-feature prompts and prototype-based logit augmentation are trained alongside an Angular Separation Loss that penalizes off-diagonal cosine similarity between class prototypes, preventing the prototype collapse that afflicts rare classes under extreme imbalance. To counteract the skewed label distribution, the training regime combines asymmetric focal loss, inverse-frequency weighted sampling, temporal Mixup, Exponential Moving Average, and per-class threshold calibration. The Biological State Machine decoder replaces naive gap merging with a physiologically grounded forward-only state transition over anatomy labels, eliminating the fragmentation artefact that produced hundreds of spurious anatomy events per video in the prior approach and reducing per-video anatomy output to 2--3 clinically realistic events. On the held-out RARE-VISION test set comprising three NaviCam examinations (161,025 frames), the updated pipeline achieves an overall temporal mAP@0.5 of 0.3597 and mAP@0.95 of 0.3399, representing a relative improvement of 46% and 44% respectively over the prior submission, with total inference completed in approximately 21 minutes on a single GPU.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑