This app uses Tensorflow Object Detection API and Tensorflow Lite for developing machine learning mobile app. The ssd_mobilenet_v2_quantized_300x300_coco model is used for transfer training with dataset from J.A. Quinn (http://air.ug/microscopy). The performance of this model is evaluated by COCO detection metrics. The scores are as follow:
Mean Average Precision (mAP) @[ IoU=0.50:0.95 ] = 0.223549
Mean Average Precision (mAP) @[ IoU=0.50 ] = 0.627176
Mean Average Precision (mAP) @[ IoU=0.75 ] = 0.085073
You can use this app only for helping Mycobacterium tuberculosis detection. It CAN'T be used for diagnosis without any supervision from medical expert.