imdb structure of caltech101
Show older comments
I want to pass arguments to @ccn_train function in matconvent for retraining alexnet on caltech 101 dataset. It requires imdb structure of caltech101. Can anyone please guide me how to create the imdb.mat (having data, labels,set,sets).
6 Comments
Anas Samara
on 5 Oct 2016
Please help in this?
José-Luis
on 5 Oct 2016
You'd probably be better off asking the developers of the matconvnet toolbox. This is not native Matlab.
Have you read the documentation on their site?
neena aloysius
on 14 Oct 2016
You're not going to find many people here that know about matconvnet AND imdb structure (whatever that means). From the lack of responses, I would guess that number of people is 0.
edit: Well, 1 as it turns out.
Walter Roberson
on 14 Oct 2016
I know a small bit about both of those, but I am overloaded and do not have the resources to take this on.
Walter Roberson
on 14 Oct 2016
There are difficulties compiling matconvnet on OS-X and difficulties compiling on Windows
I have some virtual machines running MS Windows, in which I have installed various MATLAB versions, but one thing I have not done is install any compilers -- I would have to clone off a new 140 GB virtual machine each time because the compiler installations interact with each other so in order to be able to accurately log what needs to be done for each setup, I would need the equivalent of a clean install. It also turns out that CUDA is not available to virtual machines except possibly by using a university project whose site has not been reachable when I tried. I dislike booting up into MS Windows directly, it's like driving backwards on the wrong side of the road -- possible but Not Fun.
Answers (1)
h612
on 31 Mar 2017
0 votes
Hi. Download the caltech101 dataset. This folder has .mat file with dataset variables. You can edit/use those variables in cnn training. Good luck!
Categories
Find more on Deep Learning Toolbox in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!