Foundation Model for Endoscopy Video Analysis via Large-Scale Self-supervised Pre-train

Wang, Zhao; Liu, Chang; Zhang, Shaoting; Dou, Qi

doi:10.1007/978-3-031-43996-4_10

Cited by 15 publications

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite its critical role, endoscopic research and training are hindered by the scarcity of visual data, as capturing images inside the body with small endoscopes is inherently difficult. The dire need for a diverse and quality-rich collection of clinical endoscopy videos [65] underscores the urgency for breakthroughs in medical generative AI. We aim to build a powerful endoscopy video simulator and create an extensive array of high-quality endoscopy videos, thus enriching the resources available for medical professionals [10] and improving training data for surgical robots and AI algorithms.…”

Section: Introductionmentioning

confidence: 99%

Identity of university Chinese heritage language learners in Hong Kong

Li¹,

李蓁²

View full text Add to dashboard Cite

The core of out-of-distribution (OOD) detection is to learn the in-distribution (ID) representation, which is distinguishable from OOD samples. Previous work applied recognition-based methods to learn the ID features, which tend to learn shortcuts instead of comprehensive representations. In this work, we find surprisingly that simply using reconstruction-based methods could boost the performance of OOD detection significantly. We deeply explore the main contributors of OOD detection and find that reconstructionbased pretext tasks have the potential to provide a generally applicable and efficacious prior, which benefits the model in learning intrinsic data distributions of the ID dataset. Specifically, we take Masked Image Modeling as a pretext task for our OOD detection framework (MOOD). Without bells and whistles, MOOD outperforms previous SOTA of one-class OOD detection by 5.7%, multi-class OOD detection by 3.0%, and near-distribution OOD detection by 2.1%. It even defeats the 10-shot-per-class outlier exposure OOD detection, although we do not include any OOD samples for our detection.

show abstract

Section: Introductionmentioning

confidence: 99%