Readme update

README.md

@@ -4,6 +4,11 @@
 
 The purpose of this project is to define **2D Object Recognition Tests**.
 
+## Current project status
+
+Github is used for the tracking of tasks. See repository Issues.
+
+## Task definition
 ### Object Schema
 
 Objects exist within a 2D space. The space has a width and height. Each location in the space can be identified with an X and Y integer coordinate. Each location may have a Feature.
@@ -29,13 +34,13 @@ Initially, all Features consist of a simple data type, but should be extensible
 
 The Object Library can be found in `objects/`. Each YAML file within this directory contains one object definition in the format specified in "Object Schema" above.
 
-There are currently 2 objects in the library.
+![example object with agent](doc/images/objectSpace.jpeg)
 
 ## Agency
 
 An Agent can exist in a location within an Object space. An Agent observing an object will receive features in space according to the location of its sensors.
 
-![Agent picture](https://discourse-cdn-sjc2.com/standard14/uploads/numenta/original/2X/4/49d9249b29105c9efa9eb0bbfa5b53e7f3ee369a.jpeg)
+![Agent picture](doc/images/agentSensors.jpeg)
 
 Each Agent has exactly 4 sensors:
 - North (Agent Y - 1)
@@ -44,17 +49,31 @@ Each Agent has exactly 4 sensors:
 - West  (Agent X - 1)
 
 At one time step, an Agent can be at only one location in Object space. Each sensor has access to the Feature beneath it.
+**Agents should use their sensors to attempt to identify the object under observation at each time step.**
+
+## Algorithm concept
+
+We need to build a 3-layer Network for each sensor which has an object pooling layer as described in the [Columns Paper](https://numenta.com/neuroscience-research/research-publications/papers/a-theory-of-how-columns-in-the-neocortex-enable-learning-the-structure-of-the-world/) 
+
+above a 2-layer location/sensor circuit as described in [Columns+](https://numenta.com/neuroscience-research/research-publications/papers/locations-in-the-neocortex-a-theory-of-sensorimotor-object-recognition-using-cortical-grid-cells/):
+
+![Three layer network](doc/images/ThreeLayer.jpeg)
+
+Object layers must share representations between cortical columns via lateral connections:
+![Lateral connections](doc/images/lateral.jpeg)
+
+For code examples, see the [supporting paper for the Columns paper](https://github.com/numenta/htmpapers/tree/master/frontiers/a_theory_of_how_columns_in_the_neocortex_enable_learning_the_structure_of_the_world)
 
-Agents should use their sensors to attempt to identify the object under observation at each time step.
 
-### Test Challenge \#1
+## Project structure
 
-### JavaScript
-See javascript [readme](javascript/).
+### Visualization of the objects from the Object Library
+See javascript [readme](objectVisualizer/).
 
-Code in the `javascript/` subfolder can be used to visualize Objects in the Object Library.
+### Java
+This folder now contains just "Hello World!" program in Java. Code here if you want to use Java language.
 
 ### Python
-See python [readme](python/).
+Currently used language for the development.
+See python [readme](python/) for instructions how to run.
 
-Code in the `python/` subfolder contains the beginnings of simple `Agent` and `Environment` implementations.

python/README.md

@@ -13,39 +13,27 @@ Inside the `python` directory, run:
 python -m pip install -r requirements.txt
 ```
 
+### Code style / formatting
+
+Setup your IDE to 4 spaces indentation.
+
 ## Run the tests
 
 ```
 python -m unittest tests/*.py
 ```
-# Run
+# Run experiment
 
 Just run in the python folder
 ```
 python main.py
 ```
 
-# Using visualization tool
+## Using visualization tool
 
 You can use [visualisation tool for HTM systems](https://github.com/htm-community/HTMpandaVis).
-Install what is neccessary according to project readme & enable using pandaVis in the main.py by setting appropriate flag at the beginning of the script.
-Firstly run the vis tool in terminal, then run this script in terminal. It will get connected through TCP and show state of HTM system.
-
-# Code style / formatting
-
-Project uses flake8(quality code check) and black(code formatter).
-Install them globally so your IDE can find them or point your IDE at your py environment.
+Now work in progress - TODO howto procedure
 
-```
-python -m pip install flake8
-python -m pip install black
-```
 
-Setup your IDE to 4 spaces indentation.
 
-## Manually apply formatting
-Other option is just to apply formatting to file manually.
-```
-black path/to/file.py
-```