how to apply this method to image segmentation?

[DerekRay]

spatious radius or range radius not needed?

[masqm]

Hi Derek,

1. Yes, our algorithm can be used to image segmentation.

To adopt mean-shift in image segmentation, you can refer to:
https://github.com/MichaelMansour256/image-segmentation
This a simple image segmentation example. They use mean-shift algorithm in sklearn to achieve segmentation.

If you want use our algorithm instead of it, you can replace

bandwidth = estimate_bandwidth(x, quantile=.2,n_samples=500)
ms = MeanShift(bandwidth=bandwidth, bin_seeding=True)

to our the example code shown in README.md

It is important to note that segmentation of small-size images may be slower than sklearn since our algorithm uses GPUs. However, in the large-size image segmentation, our algorithm will have a speed advantage.

2. For your second question, yes, our algorithm doesn't need pre-configured radius(bandwidth).

# Estimate bandwidth for mean shift(Select 1000 points)
bandwidth = cluster.estimate_bandwidth(X[0:999])
bandwidth_gpu = 2*bandwidth/(X.max()-X.min())

The code above can estimate a suitable bandwidth.

Best,
Masqm

[DerekRay]

Hi Derek,

1. Yes, our algorithm can be used to image segmentation.

To adopt mean-shift in image segmentation, you can refer to: https://github.com/MichaelMansour256/image-segmentation This a simple image segmentation example. They use mean-shift algorithm in sklearn to achieve segmentation.

If you want use our algorithm instead of it, you can replace

bandwidth = estimate_bandwidth(x, quantile=.2,n_samples=500)
ms = MeanShift(bandwidth=bandwidth, bin_seeding=True)

to our the example code shown in README.md

It is important to note that segmentation of small-size images may be slower than sklearn since our algorithm uses GPUs. However, in the large-size image segmentation, our algorithm will have a speed advantage.

2. For your second question, yes, our algorithm doesn't need pre-configured radius(bandwidth).

# Estimate bandwidth for mean shift(Select 1000 points)
bandwidth = cluster.estimate_bandwidth(X[0:999])
bandwidth_gpu = 2*bandwidth/(X.max()-X.min())

The code above can estimate a suitable bandwidth.

Best, Masqm

Thank you very much!