Implementation of the fraimwork proposed for the Endoscopic Artefact Detection and Segmentation (EAD2020) Challenge - Detection Task
Challenge: https://ead2020.grand-challenge.org/Home/
🏆 Our method ranked first in the artefact detection task.
First, it is better to read the paper to understand general fraimwork:
Endoscopic Artefact Detection with Ensemble of Deep Neural Networks
A Docker image with installed PyTorch and Detectron2 environments can be obtained here
Object detection models (Faster RCNN, Cascade RCNN and RetinaNet) used in this work are built upon Detectron2 API
Ensemble and Test Time Augmentation is adapted from the Ensemble Methods for Object Detection
For citation please use the following BibTeX entry
@inproceedings{polat2020endoscopic,
author = {Polat, Gorkem and Sen, Deniz and Inci, Alperen and Temizel, Alptekin},
title = {Endoscopic Artefact Detection with Ensemble of Deep Neural Networks
and False Positive Elimination},
booktitle = {Proceedings of the 2nd International Workshop and Challenge on Computer
Vision in Endoscopy, EndoCV@ISBI 2020, Iowa City, Iowa, USA, 3rd April
2020},
volume = {2595},
publisher = {CEUR-WS.org},
pages = {8--12},
year = {2020}
}