diatoms_detector/README.md

# Diatoms detector

Detects diatoms in 256 x 256 microscopic images, based on bounding box annotations. Negative samples are extracted as randomly created bounding boxes outside of the manual annotations. Uses a (the) imagenet-pretrained inceptionv3 model from `keras`.
InceptionV3 true/false classifier 2018-04-10 13:04:13 +02:00			`# Diatoms detector`

			Detects diatoms in 256 x 256 microscopic images, based on bounding box annotations. Negative samples are extracted as randomly created bounding boxes outside of the manual annotations. Uses a (the) imagenet-pretrained inceptionv3 model from `keras`.
Initial commit 2018-04-10 13:00:47 +02:00