Back to projects
Final Year Project • In ProgressComputer VisionDeep LearningMedical AIResearch

Colorectal Polyp Temporal Validation

Research-focused computer vision project investigating temporal validation of colorectal polyp detection across video/endoscopic frames, with the goal of developing a robust temporal evaluation pipeline for medical AI.

🔬 Currently working on this project and preparing a research paper.

Architecture

Endoscopic video frame sequence

Frame-level object detection & tracking

Temporal consistency evaluation layer

Temporal validation metrics pipeline

Results

Currently actively working on this project and preparing a research paper.

Developing a benchmark evaluation methodology for video-based temporal polyp validation.

Challenges

Frame-to-frame temporal inconsistency in video inference

Inter-frame motion artifacts in endoscopic camera streams

Balancing spatial precision with temporal stability across frames

Future Improvements

Completing research paper submissionComprehensive video benchmark evaluationIntegration with real-time clinical video streams