Organize and add seg model

Classification Sample Models/classification-of-HnE-stained-histological-breast-cancer-images/README.md

@@ -0,0 +1,25 @@
+## Dataset
+
+Dataset used is available at [this repository](https://rdm.inesctec.pt/dataset/nis-2017-003). 
+
+The image dataset is composed of high-resolution (2040 × 1536 pixels), uncompressed, and annotated H&E stain images from the Bioimaging 2015 breast histology classification challenge.
+
+All the images are digitized with the same acquisition conditions, with magnification of 200× and pixel size of 0.42μm × 0.42μm. Each image is labeled with one of four classes: i) normal tissue, ii) benign lesion, iii) in situ carcinoma and iv) invasive carcinoma 
+
+## Image preprocessing
+
+Patches of size 512 x 512 are extracted with a stride of 256. Before training, the pixels values are normalized and centered. 
+
+## Model & Citation
+In this notebook, there's a Keras implementation of the patch-wise network of the paper <a href="https://arxiv.org/abs/1803.04054">Two-Stage Convolutional Neural Network for Breast Cancer Histology Image Classification</a>:
+
+```
+@inproceedings{nazeri2018two,
+  title={Two-Stage Convolutional Neural Network for Breast Cancer Histology Image Classification},
+  author={Nazeri, Kamyar and Aminpour, Azad and Ebrahimi, Mehran},
+  booktitle={International Conference Image Analysis and Recognition},
+  pages={717--726},
+  year={2018},
+  organization={Springer}
+}
+```

README.md

@@ -1,5 +1,8 @@
 # caMicroscope-tfjs-models
-Sample models for caMicroscope
+Classification and segmentation sample models for caMicroscope.
 
 ## Common Issues:
 - ***Provided weight data has no target variable***: This might be a common issue for the model to fail to load. To avoid this, make sure that the keras model is loaded exactly once. To make sure, reset the runtime and load the model before carrying to conversion. This case also goes when your converting using the bash command. The Keras model which you saved must be loaded just once. To make sure, save the model, reset the runtime, load the model and save it again. Multiple loads seem to mess up the layer names. You can track this issue further [here](https://github.com/tensorflow/tfjs/issues/755).
+- Don't use Keras' Lambda in model definition. If want to use, save weights, remove Lambda calls, load weights again and then save the complete model as .hdf5
+
+Feel free to contribute:).

Segmentation Sample Models/mitosis-segmentation/README.md

@@ -0,0 +1,13 @@
+# Mitosis Segmentation
+## Dataset 
+The dataset used is available in [this](http://www.andrewjanowczyk.com/use-case-5-mitosis-detection/) post.
+
+The dataset consist of 311 images of size 2,000 x 2,000 @*40x* selected from 12 breast cancer (BCa) patients. In total there are 550 mitosic centers expertly identified using a focal microscope.
+
+This sample model is trained on br_masks for demonstration purpose to try to understand how to test models in caMicroscope.
+## Preprocessing
+Pixel values are normalized between 0-1.
+## Model
+U-net: 256 -> 128 -> 64 -> 32 -> 16 -> 32 -> 64 -> 128 -> 256.
+
+Input image size: 256 x 256 x 3