X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=blobdiff_plain;f=README.md;h=d0ec215a1767bd07bfc53b0e0758ce1891c39bd1;hb=77153ad6f6acb94a5132e9930722500cd93a6960;hp=a774d628c37ca6706d4cb269b552b57ac8edcd55;hpb=94548cd44f585308c13511b82320a04d96a0731c;p=pysvrt.git

diff --git a/README.md b/README.md
index a774d62..d0ec215 100644
--- a/README.md
+++ b/README.md
@@ -1,9 +1,44 @@
 # Introduction #
 
-This is the port of the Synthetic Visual Reasoning Test to the pytorch
-framework.
+This is a port of the Synthetic Visual Reasoning Test problems to the
+pytorch framework, with an implementation of two convolutional
+networks to solve them.
 
-The main function is
+# Installation and test #
+
+Executing
+
+```
+make -j -k
+./test-svrt.py
+```
+
+should generate an image example.png in the current directory.
+
+Note that the image generation does not take advantage of GPUs or
+multi-core, and can be as fast as 10,000 vignettes per second and as
+slow as 40 on a 4GHz i7-6700K.
+
+# Vignette generation and compression #
+
+## Vignette sets ##
+
+The svrtset.py implements the classes `VignetteSet` and
+`CompressedVignetteSet` with the following constructor
+
+```
+__init__(problem_number, nb_samples, batch_size, cuda = False, logger = None)
+```
+
+and the following method to return one batch
+
+```
+(torch.FloatTensor, torch.LongTensor) get_batch(b)
+```
+
+## Low-level functions ##
+
+The main function for genering vignettes is
 
 ```
 torch.ByteTensor svrt.generate_vignettes(int problem_number, torch.LongTensor labels)
@@ -20,7 +55,7 @@ The returned ByteTensor has three dimensions:
  * Pixel row
  * Pixel col
 
-Two functions additional functions
+The two additional functions
 
 ```
 torch.ByteStorage svrt.compress(torch.ByteStorage x)
@@ -32,27 +67,28 @@ and
 torch.ByteStorage svrt.uncompress(torch.ByteStorage x)
 ```
 
-provides a lossless compression scheme adapted to the ByteStorage of
+provide a lossless compression scheme adapted to the ByteStorage of
 the vignette ByteTensor (i.e. expecting a lot of 255s, a few 0s, and
 no other value).
 
-They allow to reduce the memory footprint by a factor ~50, and may be
-usefull to deal with very large data-sets and avoid re-generating
-images at every batch.
+This compression reduces the memory footprint by a factor ~50, and may
+be usefull to deal with very large data-sets and avoid re-generating
+images at every batch. It induces a little overhead for decompression,
+and moving from CPU to GPU memory.
 
 See vignette_set.py for a class CompressedVignetteSet using it.
 
-# Installation and test #
+# Testing convolution networks #
 
-Executing
+The file
 
 ```
-make -j -k
-./test-svrt.py
+cnn-svrt.py
 ```
 
-should generate an image example.png in the current directory.
+provides the implementation of two deep networks, and use the
+compressed vignette code to allow the training with several millions
+vignettes on a PC with 16Gb and a GPU with 8Gb.
 
-Note that the image generation does not take advantage of GPUs or
-multi-core, and can be as fast as 10,000 vignettes per second and as
-slow as 40 on a 4GHz i7-6700K.
+The networks were designed by Afroze Baqapuri during an internship at
+Idiap.