torch-contour 0.0.4

Differentiable contour to mask and contour to distance map implementation with PyTorch

torch_contour

Example of torch contour on a circle when varying the number of nodes

This library contains 2 pytorch layers for performing the diferentiable operations of :

1. contour to mask
2. contour to distance map.

It can therefore be used to transform a polygon into a binary mask or distance map in a completely differentiable way. In particular, it can be used to transform the detection task into a segmentation task. The two layers have no learnable weight, so all it does is apply a function in a derivative way.

Input (Float):

A polygon of size $2 \times n$ with
with $n$ the number of nodes

Output (Float):

A mask or distance map of size $1 \times H \times W$.
with $H$ and $W$ respectively the Heigh and Width of the distance map or mask.

Important:

The polygon must have values between 0 and 1. It is therefore important to apply a sigmoid function before the layer.*.

The predicted polygon must be ordered in counter-clockwise.

Example:

check out example.ipnb