forked from axel-angel/master-project
-
Notifications
You must be signed in to change notification settings - Fork 0
/
links
42 lines (34 loc) · 2.08 KB
/
links
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
technical blog: http://colah.github.io/
article: https://cs231n.github.io/understanding-cnn/
article: http://cs.stanford.edu/people/karpathy/cnnembed/
visualize filters: http://nbviewer.ipython.org/github/BVLC/caffe/blob/master/examples/filter_visualization.ipynb
courses on NN and CNN: https://www.youtube.com/playlist?list=PL6Xpj9I5qXYEcOhn7TqghAJ6NAPrNmUBH
train from python: http://nbviewer.ipython.org/github/BVLC/caffe/blob/tutorial/examples/01-learning-lenet.ipynb
fine-tune from python: http://nbviewer.ipython.org/github/BVLC/caffe/blob/tutorial/examples/03-fine-tuning.ipynb
caffe py layer: https://github.com/BVLC/caffe/blob/master/python/caffe/test/test_python_layer.py https://gist.github.com/shelhamer/8d9a94cf75e6fb2df221
caffe py layer params: https://github.com/BVLC/caffe/pull/2001
BSDS dataset http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/resources.html
paper CNN codes for boundary detection: http://sites.skoltech.ru/compvision/projects/n4/
caffe:
pycaffe backward() updates the weights
https://groups.google.com/forum/#!searchin/caffe-users/python$20backward/caffe-users/NKsSbZ3boGg/z1KAf3A4dT4J
How to update layer parameters from python
https://github.com/BVLC/caffe/issues/1855
servers:
http://wiki.epfl.ch/cvlab-it/resources/computing
http://wiki.epfl.ch/cvlab-servers
to get started with CNNs in general you can read this paper:
http://yann.lecun.com/exdb/publis/pdf/lecun-01a.pdf
Also there are a lot of tutorials, video lectures, etc. online
here is an example: http://deeplearning.net/tutorial/lenet.html
here a list of implementations of CNNS: http://deeplearning.net/software_links/ if you want to play with them.
More specific to the projects:
for proj 1:
the talk I told you about:
+http://techtalks.tv/talks/plenary-talk-are-deep-networks-a-solution-to-curse-of-dimensionality/60315/
and a couple of papers
http://cs.nyu.edu/~zaremba/docs/understanding.pdf
http://theorycenter.cs.uchicago.edu/REU/2014/final-papers/sauder.pdf
(for proj2:)
http://www-etud.iro.umontreal.ca/~kivinenj/papers/aistats14.pdf
http://infoscience.epfl.ch/record/203512/files/preprint_1.pdf