Nevus/ Moles Detection with SVM

Hello Everyone, I will need someone to finish my project for human nevus detection (kwdikas edge [login to view URL]) using SVM or a simple NN . You can use your datasets or use images downloaded from internet with Extreme picture finder to train it. Im sure they can help but detecting small moles on skin is difficult in general, as the parts are very small, so it is better for you to choose big images. For example, apple of the eye can be mismatch when detecting moles, due to color, although moles are smaller as i attach on [login to view URL] image.. If you think we can use face or limbs detection to first detect the human skin it would be ok but i am open to proposals on how this can be used. The export should be the mole detection. Thank you a lot in advance