The Gurus of Image Classification

Francis Bach    http://www.di.ens.fr/~fbach/

Alyosha Efros    http://www.cs.cmu.edu/~efros/

D.A. Forsyth    http://luthuli.cs.uiuc.edu/~daf/

Zaid Harchaoui    http://www.harchaoui.eu/zaid/en/

Martial Hebert    http://www.cs.cmu.edu/~hebert/

Christoph Lampert    http://pub.ist.ac.at/~chl/



Ivan Laptev    http://www.di.ens.fr/~laptev/



Aude Oliva     http://cvcl.mit.edu/aude.htm



Jean Ponce     http://www.di.ens.fr/~ponce/enindex.html


Deva Ramanan     http://www.ics.uci.edu/~dramanan/



Cordelia Schmid     http://lear.inrialpes.fr/people/schmid/



Josef Sivic     http://www.di.ens.fr/~josef/



Antonio Torralba     http://web.mit.edu/torralba/www/



Andrew zisserman    http://www.robots.ox.ac.uk/~vgg/

Tutorials collections

4. Supervised learning, SVMs, kernel methods, Francis Bach
5. Instance-level recognition (part 1), Cordelia Schmid
6. Instance-level recognition (part 2), Josef Sivic
7. Large-scale visual search (part 1), Josef Sivic
8. Large-scale visual search (part 2), Cordelia Schmid

9. Bag-of-Features models for category-level classification, Cordelia Schmid
10. Sparse coding and dictionary learning for image analysis, Jean Ponce
11. Category-level localization, Andrew Zisserman

12. Learning with structured inputs and outputs, Christoph Lampert
13. Large-scale learning, Zaid Harchaoui
14. More words and bigger pictures, David Forsyth

15. Human actions (part 1), Ivan Laptev
16. Human actions (part 2), David Forsyth
17. Human pose estimation, Deva Ramanan
18. Geometry for recognition and scene analysis, Martial Hebert

19. Human vision, Aude Olivia
20. Large scale recognition / Context / many categories, Antonio Torralba
21. Big data, Alyosha Efros

Useful Image Processing Tutorial Web Sites

Here is a short list of useful links for image processing that I have found over the last few years. Although most use Photoshop, you can still learn from them and apply the same principles using other software.

The first couple links aren't about image processing but I feel it necessary to learn much of the information contained within them to get a better understanding of Digital Photography.

http://www.cambridgeincolour.com/tutorials.htm
http://www.clarkvision.com/imagedetail/does.pixel.size.matter/

http://www.astropix.com/HTML/J_DIGIT/TOC_DIG.HTM
http://www.astronomyphotos.com/PhotoshopTechniques1.htm
http://www.astronomyphotos.com/PhotoshopTechniques2.htm
http://www.danielmccauley.com/tutorials.html
http://www.iceinspace.com.au/projects.html
http://www.geocities.com/ex_cathedra_3/Processing-Photoshop.html
http://www.budguinn.com/pages/tutorials/star%20elongation.htm

Lunar and Planetary Stacking and Processing with Registax
http://www.davesastro.co.uk/techniques/registax_tutorial/index.html
http://www.russsscope.net/staxtutorial.htm
http://www.astrophotography-tonight.com/registax-4-tutorial/
http://www.astro.shoregalaxy.com/webcam_astro.htm
http://www.scott-h.co.uk/word/?p=301
http://www.astronomie.be/registax/registax5usermanual.pdf

This last one is for a (DDP) Digital Development Processing Sharpening technique used for the processing of individual Color Channels before colorizing and combining. However I have found the technique to be useful for much more.

Increasing Heap Memory Size (Maximum allocation) in Weka

During my PhD work, i needed to use weka to testing SVM classifier for texture classification. In Weka, we can easily play with different parameters and do testing faster. Initially i was facing a problem with Heap memory allocation in Java for Weka. For large datasets it was crashing a lot and sending me a notice that Heap memory size exceeded. Normally 216 MB memory is allocated as heap memory which can be very easily increased if we know the way. I had to spend 2 days looking for this solution.  Here i would like to share it with you.

1. After the weka software being installed in the computer, we need to go to the program files where the installed files are present. for example in my case, it is \\

    C:\Program Files\Weka-3-7

    In most of the computers with Windows operating system, it is like this.

2. Then we can open the RunWeka.ini file as text file and look for the maxheap value mentioned. The above picture shows how it looks in my computer. I have changed the value from 256m to 4096m. Here we can not directly mention in GB, rather mention in MB (in this case m). Also after changing we need to save the file and restart the weka again. to save changes in any file in the windows program files we need to have the administrative account. 

3. To enable built-in Administrator Account, First you’ll need to open a command prompt in administrator mode by right-clicking and choosing “Run as administrator” (or use the Ctrl+Shift+Enter shortcut from the search box)

     Now type the following command:

     net user administrator /active:yes

4. You should see a message that the command completed successfully. Log out, and you’ll now see the     

Administrator account as a choice. (Note that the screenshots are from Vista, but this works on Windows 7). You’ll note that there’s no password for this account, so if you want to leave it enabled you should change the password.

Also to know more about Heap memory , here is a very nice link

Information on Java Heap Memory allocation

I hope this would be helpful for at least some of you. if you have any comments of suggestions please let me know by your valuable comments.