-
Notifications
You must be signed in to change notification settings - Fork 12
/
config.m
75 lines (58 loc) · 1.73 KB
/
config.m
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
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
%%%%%%%%%%%%%%%%%%%%
% The global configuration file
% Holds all settings used in all parts of LDA+MRF, enabling the exact
% reproduction of the experiment at some future date.
%%%%%
% DIRECTORIES
%%%%%
% Directory holding the experiment
RUN_DIR = [ 'Topic-Random-Field' ];
% Directory holding all the source images
IMAGE_DIR = [ 'images/' ];
% Data directory - holds all intermediate .mat files
DATA_DIR = [ 'data/' ];
allData_fname_original = [DATA_DIR,'../RNN/data/iccv09-1-allNeighborPairs_eval.mat'];
allData_fname = [DATA_DIR, 'stanford_vqed.mat'];
evalData_fname = [DATA_DIR,'iccv09-1-allNeighborPairs_train_tiny.mat'];
%'iccv09-1-allNeighborPairs_eval.mat"
%%%%
%% GLOBAL PARAMETERS
%%%%
%% Feature representation of all images
% FOLLOWING RNN FORMAT:
% the data is a cell array of fields:
% img: [H x W x 3 uint8]
% labels: [H x W double]
% segs2: [H x W double]
% feat2: [Nd x M double]
% segLabels: [Nd x 1 double]
% adj: [Nd x Nd logical]
% vq: [Nd x L ]
% Where Nd is the number of regions for data d, M is the number of
% features, L is the length of the code book.
%% model parameters savedp
model_name = [DATA_DIR, 'LDA_MRF_model.mat'];
%%%%%
%% OVERSEGMENATAION SETTINGS
%%%%%
%%%%%
%% FEATURE EXTRACTION/VQ SETTINGS
%%%%%
VQ.vocabulary_f = [DATA_DIR, 'vocabulary.mat'];
VQ.Num_Vocab = 50;
%%%%%
%% LEARNING SETTINGS
%%%%%
% Number of K neighbors in making MRF
Learn.Num_Neighbors = 4;
% How many topics in TRF
Learn.Num_Topics = 8;
% How many prototypes in a topic
Learn.Num_Prototypes = 3;
% Max number of VB-EM iterations
Learn.Max_Iterations = 40;
% Max number of VB-EM iterations for a document
Learn.V_Max_Iterations = 20;
%%%%%
%% EXPERIMENT SETTINGS
%%%%%