VinBigData X-ray Abnormalities Detection requires kagglers to localize and classify thoracic abnormalities from chest radiographs.
Radiologists face many daily challenges, especially on chest radiograph. The interpretation of chest X-rays can lead to medical misdiagnosis, even for the best practicing doctor. The current existing solution only classifies X-ray images into a list of findings, without specification of their locations on the image which sometimes leads to inexplicable results. BY localizing findings from the X-ray images, it would provide doctors with more meaningful diagnostic assistance.
Single fold VFNet trained on 1024x1024 image size, with Weighted-Box Fusion (WBF) pre-processing.
Work in progress...