X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=blobdiff_plain;f=README.md;h=4efe67fbc4538875228fdacddbb3ab38c3c9e113;hb=f2da660a51ed51cab4d81df29109e26e0da347bc;hp=44e9e1af4cc11610119e7c22d58da49d64181d61;hpb=b39a4f3da90f5af903a4d23e94e9b7823ed7d205;p=pysvrt.git

diff --git a/README.md b/README.md
index 44e9e1a..4efe67f 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,13 @@
 # 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
 
 ```
-torch.ByteTensor generate_vignettes(int problem_number, torch.LongTensor labels)
+torch.ByteTensor svrt.generate_vignettes(int problem_number, torch.LongTensor labels)
 ```
 
 where
@@ -22,8 +23,55 @@ The returned ByteTensor has three dimensions:
 
 # Installation and test #
 
+Executing
+
 ```
 make -j -k
-./built.py
 ./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 compression #
+
+The two additional functions
+
+```
+torch.ByteStorage svrt.compress(torch.ByteStorage x)
+```
+
+and
+
+```
+torch.ByteStorage svrt.uncompress(torch.ByteStorage x)
+```
+
+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).
+
+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.
+
+# Testing convolution networks #
+
+The file
+
+```
+cnn-svrt.py
+```
+
+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.
+
+The networks were designed by Afroze Baqapuri during an internship at
+Idiap.