Raabin-WBC: a large free access dataset of white blood cells from normal peripheral blood

摘要: Accurate and early detection of peripheral white blood cell anomalies plays a crucial role in the evaluation an individual9s well-being. The emergence new technologies such as artificial intelligence can be very effective achieving this. In this regard, most state-of-the-art methods use deep neural networks. Data significantly influence performance generalization power machine learning approaches, especially To that end, we collected large free available dataset cells from normal samples called Raabin-WBC. Our contains about 40000 artifacts (color spots). reassure correct data, significant number were labeled by two experts, ground truth nucleus cytoplasm extracted experts for some (about 1145), well. provide necessary diversity, various smears have been imaged. Hence, different cameras microscopes used. Raabin-WBC used tasks classification, detection, segmentation, localization. We also did primary experiments on Raabin-WBC, showed how methods, networks, was affected mentioned diversity.