English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Accurate polyp detection is essential for assisting clinical rectal cancerdiagnoses. Colonoscopy videos contain richer information than still images,making them a valuable resource for deep learning methods. Great efforts havebeen made to conduct video polyp detection through multi-frame temporal/spatialaggregation. However, unlike common fixed-camera video, the camera-moving scenein colonoscopy videos can cause rapid video jitters, leading to unstabletraining for existing video detection models. Additionally, the concealednature of some polyps and the complex background environment further hinder theperformance of existing video detectors. In this paper, we propose the\textbf{YONA} (\textbf{Y}ou \textbf{O}nly \textbf{N}eed one \textbf{A}djacentReference-frame) method, an efficient end-to-end training framework for videopolyp detection. YONA fully exploits the information of one previous adjacentframe and conducts polyp detection on the current frame without multi-framecollaborations. Specifically, for the foreground, YONA adaptively aligns thecurrent frame's channel activation patterns with its adjacent reference framesaccording to their foreground similarity. For the background, YONA conductsbackground dynamic alignment guided by inter-frame difference to eliminate theinvalid features produced by drastic spatial jitters. Moreover, YONA appliescross-frame contrastive learning during training, leveraging the ground truthbounding box to improve the model's perception of polyp and background.Quantitative and qualitative experiments on three public challenging benchmarksdemonstrate that our proposed YONA outperforms previous state-of-the-artcompetitors by a large margin in both accuracy and speed.

YONA: You Only Need One Adjacent Reference-frame for Accurate and Fast Video Polyp Detection