Hi Stéfan, I am a computer science undergrad, presently in my second year and extremely interested in machine learning and computer vision. I found the project that you mentioned, here <https://github.com/scikit-image/scikit-image/wiki/GSoC-2014>and I find it to be really interesting to work on. I have used the haar classifier for detecting faces(mostly as a black box) and I would like to understand its working so that I can work on this project, can you please let me know how I start with this? Any good algorithms in recent research papers that I have to read? or any essential prerequisites for the project? Thanks a lot in advance. Vasanth Kalingeri On Thursday, March 28, 2013 3:05:31 PM UTC+5:30, Stefan van der Walt wrote:
Hi everyone
I've been interested in getting face detection into skimage for a while. This morning, Nathan Faggian reminded me that the highly popular Viola-Jones detector is patent encumbered (yes, if you're not careful you can use patented code in packages like OpenCV). However, the following link seems to suggest that we can work around that by training our own classifier with different features:
http://rafaelmizrahi.blogspot.com/2007/02/intel-opencv-face-detection-licens...
If there's any interest in working on this, or if you already have an algorithm available, please get in touch.
Stéfan
Hi Vasanth On Tue, Feb 18, 2014 at 9:55 AM, Vasanth Kalingeri <vasanth.kalingeri@gmail.com> wrote:
I found the project that you mentioned, here and I find it to be really interesting to work on. I have used the haar classifier for detecting faces(mostly as a black box) and I would like to understand its working so that I can work on this project, can you please let me know how I start with this? Any good algorithms in recent research papers that I have to read? or any essential prerequisites for the project?
The best place to start is by reading the paper(s) by Viola & Jones, and making sure you understand their approach fully. Then, the instructions given on the project outline can be explored to train some new filters. Pre-requisites for this project are fairly general skills such as being able to a) read and understand papers independently b) apply creative thinking around algorithms and ideas to replace patent-encumbered parts of the algorithm and c) easily implement and experiment with different approaches. However, much more important is the ability to engage with the community. A good way of doing so is through the process of filing and discussing a pull request. Regards Stéfan
participants (2)
-
Stéfan van der Walt -
Vasanth Kalingeri