DLIB C++ object detection example training really slow -


i'm using dlib's object detection example , trying train using 7 images containing 14 images of bottle. these images around 200x300 pixels, although 2 larger (the 1500x2000 pixel realm). large images contain 1 example each, , although images big, bottles match same size of bottles in smaller training images. sliding window 70x240 average size of bounding boxes drew.

it has been minimizing objective function 8+ hours on windows server machine 384gb of ram running windows 8 64bit. there's no way it's supposed take long. it's still going--it's on iteration 125...

the documentation mentions training face detector on provided set of faces in "on order of 10 seconds." because i'm running in ms visual studio 2012 arguments passed debugger? when ran face detector example, took solid 30-45 minutes training--a lot more 10 seconds.

has has similar issues , know how fix it?

thanks help!

did compile in debug mode? see: http://dlib.net/faq.html#whyisdlibslow


Comments

Popular posts from this blog

javascript - AngularJS custom datepicker directive -

javascript - jQuery date picker - Disable dates after the selection from the first date picker -