This Java-based implementation of CHREST provides a graphical application and library for running a variety of CHREST models and experiments. Currently, the graphical interface supports a range of domain-types for demonstration and test purposes. Cognitive models can be developed using a range of languages.

Latest release: download

Applications

CHREST can be used in a wide number of applications, but the following application areas are specifically supported in the GUI:

  • verbal learning
  • classification
  • pattern recognition and recall
  • visual attention and memory

Model development

CHREST models can be developed using languages supported by the JVM: the manual gives examples using Java, Lisp, Python and Ruby.

  • Ruby: See the jchrest webpage for a version of CHREST bundled as a jRuby library.
  • Lisp: Examples using ABCL can be downloaded from the CHREST download link, above.
  • Python: See the installation/running instructions in demo-1.py, which uses jython.

Screenshots

Properties Learn and Recognise
Image of properties dialog Image of screen running a simple learn-and-recognise experiment
Paired Association Model View
Image of dialog training models on a paired-associate task Image of the model's long-term memory and other data
Learning Scenes Recall of Scenes
Image of dialog for training large CHREST models Image of screen showing the recall of scenes