file-metadata should be enhanced by adding and training wavelet based alogrithms. Start with Gabor filter banks and apply it to object recognition, texture and satelite images classification, text/image language identification, sound classification. If needed can be improved by using wavelets as described in the paper mentioned.
See also:
- Gabor filter banks for texture classification (code example): http://scikit-image.org/docs/dev/auto_examples/plot_gabor.html
- Literature / Paper; Support of object recognition, texture and satelite images classification, text/image language identification, sound classification: http://www.cmapx.polytechnique.fr/~yu/publications/ICPR08Final.pdf
- patch transformation: http://people.csail.mit.edu/taegsang/Documents/CVPRPatch.pdf
- "A Tutorial of Wavelet for Pattern Recognition" (Guan-Chen Pan): https://www.researchgate.net/file.PostFileLoader.html?id=54b4d915d685ccc6468b4652&assetKey=AS%3A273675431940106%401442260715381
- Chapter 15 - BLIND SOURCE SEPARATION: http://www.mit.edu/~gari/teaching/6.555/LECTURE_NOTES/ch15_bss.pdf
Possible Categories to put for the bot:
- Category:Textures
- Category:Satellite pictures
- Category:Texts, Category:Texts by language, etc. (see also T141684: Enhance file-metadata in core by adding language recognition)
- Category:Commons audio resources
- more..?
Details:
Primary mentor:
Co-mentor:
Other mentors: (optional, Phabricator username)
Skills: (python and computer vision)
Estimated project time for a senior contributor: (2-4 weeks)