Login Register

Auto-login on future visits

Forgot your password?

PRSD Studio 3.0 becomes perClass!

We are happy to announce a major release of PRSD Studio. With the new 3.0 release the software got a new name: perClass!
What's new?

Even faster execution of your classifiers

  • We significantly increased the speed of the execution runtime (see the benchmarks), and give you precision timers so that you can measure the production speed on any platform.
  • The new deployment tool 'sdrun' allows you to easily execute trained classifiers using Matlab compiler.
    Learn more

Do you work with images?

A new range of possibilities is now available for you!

With perClass you can:

  • Extract features in local image regions, which you can freely define.
  • Work with arbitrarily-shaped pixel sets.
  • Train your classifiers on regions from many images.
More in the new chapter on "Feature extraction for images"

Need to cluster your data?

Easy, with the new 'sdcluster' function you can directly cluster your data with a user-defined model. Clustering is performed per class, preserving the original labels. So you may quickly understand latent structures in your classes.
More in the new chapter on "Cluster analysis".

Get more done with improved interactive tools

  • sdscatter adds the 'show all' menu command separately for each set of labels.
  • Save time by storing the current configuration of your sdscatter filter over multiple label sets, and easily apply it to new data.
  • Efficiently visualize class distributions with the new options of sdfeatplot.