Python with libraries like numpy, scipy, scikit-learn, scikit-image It is a favorite language for machine learning shops and CV shops, To recognize the scene using these I made a Python wrapper for the GIST feature extraction library written in C.
A. It is said that the features developed by Torralba are suitable for scene recognition. Now that Deep Learning is in full swing, I'm feeling a little bit better now, but I used it this time because it's easy to try. http://people.csail.mit.edu/torralba/code/spatialenvelope/
Lear's GIST implementation C implementation of GIST features. This time I made a Python wrapper for this. http://lear.inrialpes.fr/software
lear-gist-python This is the rapper I wrote this time. Here: https://github.com/tuttieee/lear-gist-python
You can use it like this.
import gist
import numpy as np
img = ... # numpy array containing an image
descriptor = gist.extract(img)
There is also a recognition sample using the 8 scene data set (also now ... lol) published on the above A. Torralba site (http://people.csail.mit.edu/torralba/code/spatialenvelope/). ..
I wrote a C extension for Python for the first time. If you have any suggestions, please comment.
Recommended Posts