Research

Paper

TESTING March 18, 2026

Differential Attention-Augmented BiomedCLIP with Asymmetric Focal Optimization for Imbalanced Multi-Label Video Capsule Endoscopy Classification

Authors

Podakanti Satyajith Chary, Nagarajan Ganapathy

Abstract

This work presents a multi-label classification framework for video capsule endoscopy (VCE) that addresses the extreme class imbalance inherent in the Galar dataset through a combination of architectural and optimization-level strategies. Our approach modifies BiomedCLIP, a biomedical vision-language foundation model, by replacing its standard multi-head self-attention with a differential attention mechanism that computes the difference between two softmax attention maps to suppress attention noise. To counteract the skewed label distribution, where pathological findings constitute less than 0.1% of all annotated frames, a sqrt-frequency weighted sampler, asymmetric focal loss, mixup regularization, and per-class threshold optimization are employed. Temporal coherence is enforced through median-filter smoothing and gap merging prior to event-level JSON generation. On the held-out RARE-VISION test set comprising three NaviCam examinations (161,025 frames), the pipeline achieves an overall temporal mAP@0.5 of 0.2456 and mAP@0.95 of 0.2353, with total inference completed in approximately 8.6 minutes on a single GPU.

Metadata

arXiv ID: 2603.17879
Provider: ARXIV
Primary Category: cs.CV
Published: 2026-03-18
Fetched: 2026-03-19 06:01

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.17879v1</id>\n    <title>Differential Attention-Augmented BiomedCLIP with Asymmetric Focal Optimization for Imbalanced Multi-Label Video Capsule Endoscopy Classification</title>\n    <updated>2026-03-18T16:04:50Z</updated>\n    <link href='https://arxiv.org/abs/2603.17879v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.17879v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>This work presents a multi-label classification framework for video capsule endoscopy (VCE) that addresses the extreme class imbalance inherent in the Galar dataset through a combination of architectural and optimization-level strategies. Our approach modifies BiomedCLIP, a biomedical vision-language foundation model, by replacing its standard multi-head self-attention with a differential attention mechanism that computes the difference between two softmax attention maps to suppress attention noise. To counteract the skewed label distribution, where pathological findings constitute less than 0.1% of all annotated frames, a sqrt-frequency weighted sampler, asymmetric focal loss, mixup regularization, and per-class threshold optimization are employed. Temporal coherence is enforced through median-filter smoothing and gap merging prior to event-level JSON generation. On the held-out RARE-VISION test set comprising three NaviCam examinations (161,025 frames), the pipeline achieves an overall temporal mAP@0.5 of 0.2456 and mAP@0.95 of 0.2353, with total inference completed in approximately 8.6 minutes on a single GPU.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-03-18T16:04:50Z</published>\n    <arxiv:comment>9 pages, 1 figure, ICPR 2026 RARE-VISION Competition</arxiv:comment>\n    <arxiv:primary_category term='cs.CV'/>\n    <author>\n      <name>Podakanti Satyajith Chary</name>\n    </author>\n    <author>\n      <name>Nagarajan Ganapathy</name>\n    </author>\n  </entry>"
}