LiviaNET / Git / Diff of /README.md

Models:

gilberto-topp/

LiviaNET

Downloads: 2

Diff of /README.md [e9ece0] .. [c7c3a1]

Switch to unified view


...
## Some results from our paper

* Different views of a smoothed version of contours provided by our automatic segmentation system. In these images, the thalamus, caudate, putamen and pallidum are respectively depicted in yellow, cyan, red and green.

<br>
<img src="https://github.com/josedolz/LiviaNET/blob/master/Images/NeuroRes2.jpg?raw=true" />
<br>

* Feature map activations in all convolutional layers of the FCNN (right), obtained for a given patch of the input MRI image (left). Each column corresponds to a different convolutional layer, from shallow to deeper, and each image in a row to a features map activation randomly selected in the layer.
<br>
<img src="https://github.com/josedolz/LiviaNET/blob/master/Images/NeuroFeatMaps.jpg?raw=true" />
<br>



### Known problems

	a/README.md		b/README.md
	...		...
90	## Some results from our paper	90	## Some results from our paper
91		91
92	* Different views of a smoothed version of contours provided by our automatic segmentation system. In these images, the thalamus, caudate, putamen and pallidum are respectively depicted in yellow, cyan, red and green.	92	* Different views of a smoothed version of contours provided by our automatic segmentation system. In these images, the thalamus, caudate, putamen and pallidum are respectively depicted in yellow, cyan, red and green.
93		93
94	<br>	94	<br>
95	<img src="https://github.com/josedolz/LiviaNET/blob/master/Images/NeuroRes2.jpg?format=raw" />	95	<img src="https://github.com/josedolz/LiviaNET/blob/master/Images/NeuroRes2.jpg?raw=true" />
96	<br>	96	<br>
97		97
98	* Feature map activations in all convolutional layers of the FCNN (right), obtained for a given patch of the input MRI image (left). Each column corresponds to a different convolutional layer, from shallow to deeper, and each image in a row to a features map activation randomly selected in the layer.	98	* Feature map activations in all convolutional layers of the FCNN (right), obtained for a given patch of the input MRI image (left). Each column corresponds to a different convolutional layer, from shallow to deeper, and each image in a row to a features map activation randomly selected in the layer.
99	<br>	99	<br>
100	<img src="https://github.com/josedolz/LiviaNET/blob/master/Images/NeuroFeatMaps.jpg?format=raw" />	100	<img src="https://github.com/josedolz/LiviaNET/blob/master/Images/NeuroFeatMaps.jpg?raw=true" />
101	<br>	101	<br>
102		102
103		103
104		104
105	### Known problems	105	### Known problems