stem/AI/Neural Networks/CNN/FCN/FCN.md

Fully [Convolution](../../../../Signal%20Proc/Convolution.md)al Network

[[Convolutional Layer|Convolutional]] and [[UpConv|up-convolutional layers]] with [[Activation Functions#ReLu|ReLu]] but no others (pooling)
- All some sort of Encoder-Decoder

Contractive → [UpConv](../UpConv.md)

# Image Segmentation
- For visual output
	- Previously image $\rightarrow$ vector
- Additional layers to up-sample representation to an image
	- Up-[convolution](../../../../Signal%20Proc/Convolution.md)al
	- De-[convolution](../../../../Signal%20Proc/Convolution.md)al

![fcn-uses](../../../../img/fcn-uses.png)
![fcn-arch](../../../../img/fcn-arch.png)

# Training
- Rarely from scratch
- Pre-trained weights
- Replace final layers
	- [[MLP|FC]] layers
	- White-noise initialised
- Add [UpConv](../UpConv.md) layer(s)
	- Fine-tune train
	- Freeze others
	- Annotated GT images
- Can use summed per-pixel log [[Deep Learning#Loss Function|loss]]

# Evaluation
![fcn-eval](../../../../img/fcn-eval.png)
- SDS
	- Classical method
	- 52% mAP
- FCN
	- 62% mAP
- Intersection over Union
	- IOU
	- Jaccard
	- Averaged over all images
	- $J(A,B)=\frac{|A\cap B|}{|A\cup B|}$