Update README

README.md

@@ -1,4 +1,4 @@
-This project is both an exercise and an attempt at using a webcam to predict the mouse pointer position.
+This project is mainly an exercise but also an attempt at using a webcam to predict the mouse pointer position.
 This is ( not ) achieved ( but not that far ) using a ( rather simple ) neural net.
 As a neural net needs to be trained, training data is needed.
 This project presents three python scripts: one to acquire data, one to define and train a neural net, one to test the results.
@@ -17,16 +17,16 @@ Displays a black image that covers all your screen.
 A blue pixel surrounded by a grey square is displayed ( for each new position ).
 While looking at the blue pixel, press `space` to acquire some images ( meanwhile the grey square is removed ).
 Once a point data has been gathered a new point is selected and displayed.
-Points are sampled on a 10x10 grid.
+Points are first sampled on a 10x10 grid, then on 40 locations along the screen border.
 Once the 10x10 grid has been scanned the program stops.
 Press `q` or `escape` to stop early, but a complete run shouldn't be too long.
 
 To provide some robustness to sensor noise and blinking, five images are taken for each sampled point.
-As a complete acquisition run scans 10x10 positions, it produces 500 training data.
-It usually takes between 1.5mn and 2mn for me.
+As a complete acquisition run scans 10x10 + 40 positions, it produces 700 training data.
+It usually takes about 2mn for me.
 
-You should probably aim for at least 20 000 images ( therefore performing 40 acquisition runs ).
-Perform different acquisition runs preferably at different times of the day, at different positions in front of your screen, etc...
+You should probably aim for at least 20 000 images ( therefore performing 30 acquisition runs ).
+Perform different acquisition runs preferably at different: times of the day, positions in front of your screen, screen height, screen inclination, etc...
 In short: try to make the training data representative of the final use cases.
 
 Quirk: the net is prone to learn to only use the head orientation ( thus ignoring eyes ).
@@ -46,7 +46,7 @@ File: `Training/train.py`
 The net is defined in `Training/gaze_net.py`
 To retain enough details, images are downscaled only to a 320x160 resolution.
 This is quite big, therefore training is quite long.
-With an nvidia gtx970 card, training on about 40 000 images takes about 10h.
+With an nvidia gtx970 card, training on about 40 000 images takes about 18h.
 During training the parameters are regularly saved to disk.
 Resulting files have the names: `gn_320x240_mse0.XXXX_epochY.pt`.
 In these names `mse0.XXXX` is the mean squared error for the net on the training data, and `epochY` simply is the epoch number.