Motion-Based Weak Supervision for Video Parsing with Application to Colonoscopy

TL;DR

This paper introduces an unsupervised method that uses motion cues to segment videos and weakly supervise appearance classifiers, effectively parsing colonoscopy videos into phases without extensive labeled data.

Contribution

The novel two-stage approach leverages motion cues for segmentation and weak supervision for appearance classification in medical video analysis.

Findings

Effective phase detection in colonoscopy videos

Unsupervised segmentation improves classification accuracy

Method reduces need for manual annotations

Abstract

We propose a two-stage unsupervised approach for parsing videos into phases. We use motion cues to divide the video into coarse segments. Noisy segment labels are then used to weakly supervise an appearance-based classifier. We show the effectiveness of the method for phase detection in colonoscopy videos.