plateform stringclasses 1
value | repo_name stringlengths 13 113 | name stringlengths 3 74 | ext stringclasses 1
value | path stringlengths 12 229 | size int64 23 843k | source_encoding stringclasses 9
values | md5 stringlengths 32 32 | text stringlengths 23 843k |
|---|---|---|---|---|---|---|---|---|
github | mcv-m1-project/Team5-master | TrafficSignDetection_test.m | .m | Team5-master/week4/TrafficSignDetection/TrafficSignDetection_test.m | 3,876 | utf_8 | 4a32b363bcfd524bfc179d2b7873fa7a | %
% Template example for using on the test set (no annotations).
%
function TrafficSignDetection_validation(input_dir, output_dir, pixel_method, window_method, decision_method)
% TrafficSignDetection
% Perform detection of Traffic signs on images. Detection is performed first at the pixel level
% using a... |
github | mcv-m1-project/Team5-master | colorspace_demo.m | .m | Team5-master/week4/colorspace/colorspace_demo.m | 6,856 | utf_8 | f7d66bc3e0e1bf1611fbd525c617323c | function colorspace_demo(Cmd)
% Demo for colorspace.m - 3D visualizations of various color spaces
% Pascal Getreuer 2006
if nargin == 0
% Create a figure with a drop-down menu
figure('Color',[1,1,1]);
h = uicontrol('Style','popup','Position',[15,10,90,21],...
'BackgroundColor',[1,1,1],'Value',2,...
... |
github | mcv-m1-project/Team5-master | colorspace.m | .m | Team5-master/week4/colorspace/colorspace.m | 16,178 | utf_8 | 2ca0aee9ae4d0f5c12a7028c45ef2b8d | function varargout = colorspace(Conversion,varargin)
%COLORSPACE Transform a color image between color representations.
% B = COLORSPACE(S,A) transforms the color representation of image A
% where S is a string specifying the conversion. The input array A
% should be a real full double array of size Mx3 or MxN... |
github | mcv-m1-project/Team5-master | rgb2yuv.m | .m | Team5-master/week4/color_segmentation/Other color spaces/rgb2yuv.m | 643 | utf_8 | a08cf3a0f2c43a09cc25d8c521506f7e | %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%This function rgb2yuv converts the RB matrix of an image to an YUV format%
%matrix for the image. It plots the images, if Plot Flag is eqaul to 1. %
% Code from: Mathworks
%%%%%%%%%%... |
github | mcv-m1-project/Team5-master | Template.m | .m | Team5-master/week4/Template_Matching/Manual_Templates/Template.m | 750 | utf_8 | 0020946e309b799695ef15d872f69bff | %Task 1 w4
function template = Template(signType,siz)
switch signType
case 1
template = template_model1(siz);
template = edge(template,'canny');
% figure();
% imshow(template);
case 2
template = template_model2(siz);
... |
github | mcv-m1-project/Team5-master | TrafficSignDetection.m | .m | Team5-master/week2/TrafficSignDetection.m | 6,423 | utf_8 | fb3ab5108eed95507bf253c519d06025 | %
% Template example for using on the validation set.
%
function TrafficSignDetection(directory, pixel_method, window_method, decision_method)
% TrafficSignDetection
% Perform detection of Traffic signs on images. Detection is performed first at the pixel level
% using a color segmentation. Then, using t... |
github | mcv-m1-project/Team5-master | TrafficSignDetection_test.m | .m | Team5-master/week2/TrafficSignDetection_test.m | 3,876 | utf_8 | 4a32b363bcfd524bfc179d2b7873fa7a | %
% Template example for using on the test set (no annotations).
%
function TrafficSignDetection_validation(input_dir, output_dir, pixel_method, window_method, decision_method)
% TrafficSignDetection
% Perform detection of Traffic signs on images. Detection is performed first at the pixel level
% using a... |
github | mcv-m1-project/Team5-master | mystrel.m | .m | Team5-master/week2/Morphologic_operators/mystrel.m | 1,337 | utf_8 | a8cc8a5733e4559fc8ea3365a9f20396 | % Task 1. mystrel is a function that creates the morphological structuring
% element that will be used by the morphological operators. The function
% has the next three input parameters (IP) and one output parameter (OP).
%
% size1: (IP) parameter used to create the circle and square structuring element
% size2: (IP) p... |
github | mcv-m1-project/Team5-master | colorspace_demo.m | .m | Team5-master/week2/colorspace/colorspace_demo.m | 6,856 | utf_8 | f7d66bc3e0e1bf1611fbd525c617323c | function colorspace_demo(Cmd)
% Demo for colorspace.m - 3D visualizations of various color spaces
% Pascal Getreuer 2006
if nargin == 0
% Create a figure with a drop-down menu
figure('Color',[1,1,1]);
h = uicontrol('Style','popup','Position',[15,10,90,21],...
'BackgroundColor',[1,1,1],'Value',2,...
... |
github | mcv-m1-project/Team5-master | colorspace.m | .m | Team5-master/week2/colorspace/colorspace.m | 16,178 | utf_8 | 2ca0aee9ae4d0f5c12a7028c45ef2b8d | function varargout = colorspace(Conversion,varargin)
%COLORSPACE Transform a color image between color representations.
% B = COLORSPACE(S,A) transforms the color representation of image A
% where S is a string specifying the conversion. The input array A
% should be a real full double array of size Mx3 or MxN... |
github | mcv-m1-project/Team5-master | rgb2yuv.m | .m | Team5-master/week2/color_segmentation/rgb2yuv.m | 643 | utf_8 | a08cf3a0f2c43a09cc25d8c521506f7e | %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%This function rgb2yuv converts the RB matrix of an image to an YUV format%
%matrix for the image. It plots the images, if Plot Flag is eqaul to 1. %
% Code from: Mathworks
%%%%%%%%%%... |
github | mcv-m1-project/Team5-master | TrafficSignDetection.m | .m | Team5-master/week3/TrafficSignDetection.m | 6,423 | utf_8 | fb3ab5108eed95507bf253c519d06025 | %
% Template example for using on the validation set.
%
function TrafficSignDetection(directory, pixel_method, window_method, decision_method)
% TrafficSignDetection
% Perform detection of Traffic signs on images. Detection is performed first at the pixel level
% using a color segmentation. Then, using t... |
github | mcv-m1-project/Team5-master | TrafficSignDetection_test.m | .m | Team5-master/week3/TrafficSignDetection_test.m | 3,876 | utf_8 | 4a32b363bcfd524bfc179d2b7873fa7a | %
% Template example for using on the test set (no annotations).
%
function TrafficSignDetection_validation(input_dir, output_dir, pixel_method, window_method, decision_method)
% TrafficSignDetection
% Perform detection of Traffic signs on images. Detection is performed first at the pixel level
% using a... |
github | mcv-m1-project/Team5-master | mystrel.m | .m | Team5-master/week3/Morphologic_operators/mystrel.m | 1,337 | utf_8 | a8cc8a5733e4559fc8ea3365a9f20396 | % Task 1. mystrel is a function that creates the morphological structuring
% element that will be used by the morphological operators. The function
% has the next three input parameters (IP) and one output parameter (OP).
%
% size1: (IP) parameter used to create the circle and square structuring element
% size2: (IP) p... |
github | mcv-m1-project/Team5-master | colorspace_demo.m | .m | Team5-master/week3/colorspace/colorspace_demo.m | 6,856 | utf_8 | f7d66bc3e0e1bf1611fbd525c617323c | function colorspace_demo(Cmd)
% Demo for colorspace.m - 3D visualizations of various color spaces
% Pascal Getreuer 2006
if nargin == 0
% Create a figure with a drop-down menu
figure('Color',[1,1,1]);
h = uicontrol('Style','popup','Position',[15,10,90,21],...
'BackgroundColor',[1,1,1],'Value',2,...
... |
github | mcv-m1-project/Team5-master | colorspace.m | .m | Team5-master/week3/colorspace/colorspace.m | 16,178 | utf_8 | 2ca0aee9ae4d0f5c12a7028c45ef2b8d | function varargout = colorspace(Conversion,varargin)
%COLORSPACE Transform a color image between color representations.
% B = COLORSPACE(S,A) transforms the color representation of image A
% where S is a string specifying the conversion. The input array A
% should be a real full double array of size Mx3 or MxN... |
github | mcv-m1-project/Team5-master | rgb2yuv.m | .m | Team5-master/week3/color_segmentation/rgb2yuv.m | 643 | utf_8 | a08cf3a0f2c43a09cc25d8c521506f7e | %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%This function rgb2yuv converts the RB matrix of an image to an YUV format%
%matrix for the image. It plots the images, if Plot Flag is eqaul to 1. %
% Code from: Mathworks
%%%%%%%%%%... |
github | jdiedrichsen/pcm_toolbox-master | minimize.m | .m | pcm_toolbox-master/minimize.m | 11,205 | utf_8 | 96fbfc58ee0212ebd45ca4aa2d8d8609 | function [X, fX, i] = minimize(X, f, length, varargin)
% Minimize a differentiable multivariate function using conjugate gradients.
%
% Usage: [X, fX, i] = minimize(X, f, length, P1, P2, P3, ... )
%
% X initial guess; may be of any type, including struct and cell array
% f the name or pointer to the funct... |
github | jdiedrichsen/pcm_toolbox-master | pcm_fitModelIndividCrossval.m | .m | pcm_toolbox-master/pcm_fitModelIndividCrossval.m | 13,612 | utf_8 | c241d11de9ed8cea6eeb977fae2d073f | function [T,DD,theta_hat,theta0]=pcm_fitModelIndividCrossval(Y,M,partitionVec,conditionVec,varargin);
% function [T,D,theta_hat]=pcm_fitModelIndividCrossval(Y,M,partitionVec,conditionVec,varargin);
% Fits pattern component model(s) specified in M to data from one (or more)
% subjects individually, using leave-one out c... |
github | jdiedrichsen/pcm_toolbox-master | pcm_EM_free.m | .m | pcm_toolbox-master/pcm_EM_free.m | 10,318 | utf_8 | fb685a95ebd8512717f1269e98f29dcc | function [G,h,u,l,n,jumpI,a]=pcm_EM_free(y,Z,varargin)
% estimate random-effects variance component coefficients
%
% Usage: [G,h,u,l,n,jumpI,a]=pcm_EM_free(y,Z,varargin);
%
% Estimates the variance coefficients of the model described in:
% Laird, Lange & Stram (1987).
% y_n = X b_n + Z u_n + e,
% u ~ (b... |
github | jdiedrichsen/pcm_toolbox-master | pcm_recipe_feature.m | .m | pcm_toolbox-master/recipe_feature/pcm_recipe_feature.m | 5,606 | utf_8 | 96043f5e88461cda1d2639175facc8ac | function varargout=pcm_recipe_feature
% This example uses a PCM feature model to determine correspondence of activity
% patterns cross two conditions.
% The data contains patterns for movements of 5 fingers for the left and
% right hand, respectively.
% The data come from the 12 hemispheres studied for the finger mo... |
github | estradjm/Code-Portfolio-master | svm.m | .m | Code-Portfolio-master/Machine_Learning/Speaker_Recognition/svm.m | 5,720 | utf_8 | 7023245ccf77eba26b761f29daf2bef8 | %function [EER]=SVM(mainDir)
%Support Vector Machine with Linear Kernel
%[~, folders, ~, ~,~,...
% ~, ~, ~,...
% ~, ~, ~, ~, ~,...
% ~, ~, ~,...
% ~, SvmExe, SVMtrainConfig, ...
% SVMtestConfig, ~, ~]=getALIZEpath(mainDir);
mainDir='/home/jenniffer/Desktop/data/Comb_Room_Mic/';
eval(['cd ' mainDir '/'... |
github | estradjm/Code-Portfolio-master | main.m | .m | Code-Portfolio-master/Machine_Learning/Speaker_Recognition/main.m | 11,839 | utf_8 | d422f2ea76d55dfda94902283932a130 | clear all; close all; clc;
% --------------------------------------------------------------------
% Run-Time Options
% --------------------------------------------------------------------
%
% User Configurable Options:
clean_setup=1; % Clean out all folders
deltaTestEnable=0; % Energy Derivati... |
github | old-NWTC/HydroDyn-master | ImportFastOut.m | .m | HydroDyn-master/Utilities/Simulation Toolbox/FAST2MATLAB/ImportFastOut.m | 495 | utf_8 | 8351b10223640304d3aa014db568b8a9 | %ImportFastOut
%Paul Fleming
%2/6/2013
%Quick tool for importing a FAST output file
function ImportFastOut(filename)
if nargin < 1
listing = dir('*.out');
filename = listing.name;
end
fprintf('Importing %s\n',filename);
A = importdata(filename, '\t', 8);
channels = strtrim(rege... |
github | old-NWTC/HydroDyn-master | HD2Matlab.m | .m | HydroDyn-master/Utilities/Simulation Toolbox/FAST2MATLAB/HD2Matlab.m | 16,125 | utf_8 | f9370af1f7657b20eb371cd6577a21ca | function DataOut = HD2Matlab(HD_file,hdrLines,DataOut)
%% HD2Matlab
% Function for reading HydroDyn input files in to a MATLAB struct.
%
%
%This function returns a structure DataOut, which contains the following
% cell arrays:
%.Val An array of values
%.Label An array of matching labels... |
github | old-NWTC/HydroDyn-master | Matlab2HD.m | .m | HydroDyn-master/Utilities/Simulation Toolbox/MATLAB2FAST/Matlab2HD.m | 21,687 | utf_8 | d1b8143508bd4f9e5274665e433affef | % Matlab2HD
% Function for creating a new HD file given:
% 1) An input template file
% 2) A HD parameter structure
%
% In: HDPar - A HD parameter list
% TemplateFile - A .dat file to use as a template
% OutputFilename - Desired filename of output .fst file
%
% Paul Fleming, JUNE... |
github | HarvardAgileRoboticsLab/unscented-dynamic-programming-master | demo_airplane.m | .m | unscented-dynamic-programming-master/demo_airplane.m | 8,296 | utf_8 | db0908002f464bde0869cfab4720122c | function [xtraj,uhist,v] = demo_airplane
% Airplane barrel roll demo.
clc;
close all;
dt = .025;% time step
dynamics = @yak_dynamics_mex;
%dynamics = @yak_dynamics;
% initial conditions:
x0 = [-3 0 1.5 0.997156 0 0.075366 5 0 0 0 0 0]';
utrim = [41.6666 106 74.6519 106]';
% final conditions:
xg = [3 ... |
github | HarvardAgileRoboticsLab/unscented-dynamic-programming-master | UDP.m | .m | unscented-dynamic-programming-master/UDP.m | 27,394 | utf_8 | b3c4f825b55341510150b9f6a9a6e79b | function [x, u, L, Vx, Vxx, cost, trace, stop] = UDP(DYNCST, x0, u0, Op, dynamics, dt, scale)
% UDP - solve the deterministic finite-horizon optimal control problem.
%
% minimize sum_i CST(x(:,i),u(:,i)) + CST(x(:,end))
% u
% s.t. x(:,i+1) = DYN(x(:,i),u(:,i))
%
%
% This code is adapte... |
github | HarvardAgileRoboticsLab/unscented-dynamic-programming-master | demo_pendulum.m | .m | unscented-dynamic-programming-master/demo_pendulum.m | 8,721 | utf_8 | 0aeee5cc65566900d0ccd4225ccead37 | function demo_pendulum
% A pendulum swing-up demo.
clc;
dt = .1;% time step
% quadratic costs
Qf = 30*eye(2);
Q = .3*eye(2);
R = .3;
% optimization problem
T = 5/dt; % horizon
x0 = [0 0]';
xg = [pi 0]';
u0 = zeros(1,T); % initial controls
% run the optimization
Op.parallel = false;
Op.tolFun = 1... |
github | HarvardAgileRoboticsLab/unscented-dynamic-programming-master | yak_dynamics.m | .m | unscented-dynamic-programming-master/yak_dynamics.m | 6,517 | utf_8 | 04146ff4beaf0c231d4d4c5c7a940f98 | function xdot = yak_dynamics(t,x,u)
%State vector:
%r = x(1:3); %Lab-frame position vector
r = x(4:6); %MRP rotation from body to lab frame
v = x(7:9); %Lab-frame velocity vector
w = x(10:12); %Body-frame angular velocity
Q = mrptodcm(r);
%Control input:
thr = u(1); %Throttle comm... |
github | HarvardAgileRoboticsLab/unscented-dynamic-programming-master | demo_cartpole.m | .m | unscented-dynamic-programming-master/demo_cartpole.m | 7,695 | utf_8 | 2d2f327fb295e6288812a622b2d85670 | function demo_cartpole
% A cart pole swing-up demo.
clc;
close all;
dt = .1;% time step
% quadratic costs
Qf = 1000*eye(4);
Q = .1*eye(4);
R = .01;
% optimization problem
T = 5/dt; % horizon
x0 = [0 0 0 0]';
xg = [0 pi 0 0]';
u0 = zeros(1,T); % initial controls
% run the optimization
Op.parallel... |
github | HarvardAgileRoboticsLab/unscented-dynamic-programming-master | iLQG.m | .m | unscented-dynamic-programming-master/iLQG.m | 24,686 | utf_8 | d641b981e85810b4331789d60341d38b | function [x, u, L, Vx, Vxx, cost, trace, stop] = iLQG(DYNCST, x0, u0, Op)
% iLQG - solve the deterministic finite-horizon optimal control problem.
%
% minimize sum_i CST(x(:,i),u(:,i)) + CST(x(:,end))
% u
% s.t. x(:,i+1) = DYN(x(:,i),u(:,i))
%
% Inputs
% ======
% DYNCST - A combined d... |
github | changjenyin/wifi_vision-master | write_sv_to_file.m | .m | wifi_vision-master/parse_csi/src/write_sv_to_file.m | 727 | utf_8 | 10809072081697bef79685efc2eba4e7 | function write_sv_to_file(svs, names, output_dir)
action_dic = containers.Map;
action_key = {'still', 'jump', 'pickbox', 'run', 'swing', 'walk'};
action_val = {0, 1, 2, 3, 4, 5};
for i = 1:length(action_key)
action_dic(action_key{i}) = action_val{i};
end
keys(action_dic);
if ~exist(output_dir, 'dir')
mkd... |
github | changjenyin/wifi_vision-master | low_avg.m | .m | wifi_vision-master/parse_csi/src/low_avg.m | 3,543 | utf_8 | e3b2ea6c1f8e230ee9841b9eb8984943 | function low_avg(folder, output_folder, width, height, NUM_ANT_PAIR)
if ~exist('NUM_ANT_PAIR', 'var')
NUM_ANT_PAIR = 4;
end
SUB_CNT = NUM_ANT_PAIR*30;
time = 5;
target_freq = 50;
avg_window_ms = 300;
parts = strsplit(folder, '/');
dirlist = dir(folder);
SV = []; names = {};
vmins = {[],[],[],[... |
github | changjenyin/wifi_vision-master | pca_denoise.m | .m | wifi_vision-master/parse_csi/src/pca_denoise.m | 1,192 | utf_8 | 77da43178d115cc93fe7b220003ddcf1 | %{
t x d -> t x 5
[data] [00000]
[data] [00000]
[data] [00000]
. .
. .
. .
[data] [00000]
(implement by pca())
[coeff, score, latent] = pca(H, 'VariableWeights', 'variance');
denoised_H = score(:, start_component, end_component);
%}
function denoised_H = PCA_denoise... |
github | changjenyin/wifi_vision-master | none.m | .m | wifi_vision-master/parse_csi/src/none.m | 3,504 | utf_8 | 5f93cbff1b01c0e7e44913ed053dde72 | function low_avg(folder, output_folder, width, height, NUM_ANT_PAIR)
if ~exist('NUM_ANT_PAIR', 'var')
NUM_ANT_PAIR = 4;
end
SUB_CNT = NUM_ANT_PAIR*30;
time = 5;
target_freq = 50;
avg_window_ms = 300;
parts = strsplit(folder, '/');
dirlist = dir(folder);
SV = []; names = {};
vmins = {[],[],[],[... |
github | chetanborse007/EyeDetection-master | eye_detection.m | .m | EyeDetection-master/eye_detection.m | 11,386 | utf_8 | bdc57705ff1c6bc3e93f0bbabebdb9a1 |
function [left_x, right_x, left_y, right_y] = eye_detection(img)
% DESCRIPTION: Algorithm to detect Left and Right Eye.
% INPUT: RGB image
% OUTPUT: X and Y coordinates of Left and Right Eye.
% Create a Pattern Recognition Network and train it.
% [Optional] Test trained Pattern Recognition... |
github | ganggit/mmdpnc-master | ars.m | .m | mmdpnc-master/ars.m | 3,547 | utf_8 | 06964a61cc698c4bca29186492dc624b | function samples = ars(logpdf, pdfargs, N, xi, support)
% Perform adaptive rejection sampling as described in gilks & wild
% '92, and wild & gilks 93. The PDF must be log-concave. Draw N
% samples from the pdf passed in as a function handle to its log. The
% log could be offset by an additive constant, correspondin... |
github | ganggit/mmdpnc-master | handleRemovedClasses_mmc.m | .m | mmdpnc-master/handleRemovedClasses_mmc.m | 857 | utf_8 | 88bbf4dd2da4b9918788e1b1b1b5cc2e | % Copyright (C) 2007 Jacob Eisenstein: jacobe at mit dot edu
% distributable under GPL, see README.txt
function params = handleRemovedClasses_mmc(params)
%params = handleRemovedClasses(params)
if (~isfield(params,'num_fixed') || params.num_fixed == inf)
idxs = find(params(end).counts == 0);
for ctr = idxs
... |
github | ganggit/mmdpnc-master | hidupdateweight_mmc.m | .m | mmdpnc-master/hidupdateweight_mmc.m | 498 | utf_8 | a7fc1caecfb6b032bef4f10fd1eff4ea |
function [params] = hidupdateweight_mmc(params, new_class, data)
alpha = 0.01;
if nargin < 4
flag = false;
end
%must add iteratively because of the cholesky update function
for i = 1:size(data,1)
params.counts(new_class) = params.counts(new_class) - 1;
params.sums(new_class,:) = params.sums(new_class,:) -... |
github | ganggit/mmdpnc-master | hidupdateweight.m | .m | mmdpnc-master/hidupdateweight.m | 1,550 | utf_8 | 42efb9d4e87eceb09acdf76d7357f7ef |
function [params] = hidupdateweight(params, new_class, data)
alpha = 0.01;
if nargin < 4
flag = false;
end
%must add iteratively because of the cholesky update function
for i = 1:size(data,1)
params.counts(new_class) = params.counts(new_class) - 1;
params.sums(new_class,:) = params.sums(new_class,:) - dat... |
github | ganggit/mmdpnc-master | handleRemovedClasses.m | .m | mmdpnc-master/handleRemovedClasses.m | 843 | utf_8 | c86deacac3c94d2b3187644cffd6becd | % Copyright (C) 2007 Jacob Eisenstein: jacobe at mit dot edu
% distributable under GPL, see README.txt
function params = handleRemovedClasses(params)
%params = handleRemovedClasses(params)
if (~isfield(params,'num_fixed') || params.num_fixed == inf)
idxs = find(params(end).counts == 0);
for ctr = idxs
%red... |
github | ganggit/mmdpnc-master | dpmm_mmc.m | .m | mmdpnc-master/dpmm_mmc.m | 8,577 | utf_8 | 2e98d360562c8873c489d2d8a1b485c8 | % Copyright (C) 2013 Gang Chen, gangchen@buffalo.edu
% distributable under GPL, see README.txt
function [params, tElapsed] = dpmm_mmc(data, num_its, alpha, params)
%function params = dpmm(data, num_its, params)
%standard dirichlet process mixture model, with gaussian observations
%"rao-blackwellised" from, which... |
github | ganggit/mmdpnc-master | addNewClass_mmc.m | .m | mmdpnc-master/addNewClass_mmc.m | 2,460 | utf_8 | 0f177173b7d28538b1f442ecaa7a558e | % Copyright (C) 2007 Jacob Eisenstein: jacobe at mit dot edu
% distributable under GPL, see README.txt
function params = addNewClass_mmc(params, flag)
%function params = addNewClass(params)
%adds a new, empty class to the dpmm
if nargin <2
flag = false;
end
newclassidx = params.num_classes+1;
par... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | AngleWrap.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/SLAM/AngleWrap.m | 173 | utf_8 | 6e4585850dc57bd5fbce3fd38a1a12da | %% Function to handle the transformation of angle
function angle = AngleWrap(angle)
if(angle>pi)
angle=angle-2*pi;
elseif(angle<-pi)
angle = angle+2*pi;
end; |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | tinv.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/SLAM/tinv.m | 433 | utf_8 | bc8bf0fa4d4dc647e88ed009cc55657a | function tba=tinv(tab)
tba = zeros(size(tab));
for t=1:3:size(tab,1),
tba(t:t+2) = tinv1(tab(t:t+2));
end
end
%-------------------------------------------------------
function tba=tinv1(tab)
%
% calculates the inverse of one transformations
%-------------------------------------------------------
s = ... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | PlotEllipse.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/SLAM/PlotEllipse.m | 381 | utf_8 | 75ead0779e18e88a48b625e1d3e376f5 | %%%%%%%% Cavariance elipse ( Eig values) %%%%%%%%
function eH = PlotEllipse(x,P,nSigma)
eH = [];
P = P(1:2,1:2); % we are only interested in x and y part
x = x(1:2);
if(~any(diag(P)==0))
[V,D] = eig(P);
y = nSigma*[cos(0:0.1:2*pi);sin(0:0.1:2*pi)];
el = V*sqrtm(D)*y;
el = [el el(:,1)]+repmat(x,... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | DrawRobot.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/SLAM/DrawRobot.m | 506 | utf_8 | a3ec6aa3dde44aa6ae5ae373e90d5cf4 | %-------- Drawing Vehicle -----%
function DrawRobot(Xr,col,ShiftTheta);
p=0.02; % percentage of axes size
a=axis;
l1=(a(2)-a(1))*p;
l2=(a(4)-a(3))*p;
P=[-1 1 0 -1; -1 -1 3 -1];%basic triangle
theta = Xr(3)-pi/2+ShiftTheta;%rotate to point along x axis (theta = 0)
c=cos(theta);
s=sin(theta);
P=[c -s; s c]*P... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | AngleWrap.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Localisation/AngleWrap.m | 173 | utf_8 | 6e4585850dc57bd5fbce3fd38a1a12da | %% Function to handle the transformation of angle
function angle = AngleWrap(angle)
if(angle>pi)
angle=angle-2*pi;
elseif(angle<-pi)
angle = angle+2*pi;
end; |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | tinv.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Localisation/tinv.m | 433 | utf_8 | bc8bf0fa4d4dc647e88ed009cc55657a | function tba=tinv(tab)
tba = zeros(size(tab));
for t=1:3:size(tab,1),
tba(t:t+2) = tinv1(tab(t:t+2));
end
end
%-------------------------------------------------------
function tba=tinv1(tab)
%
% calculates the inverse of one transformations
%-------------------------------------------------------
s = ... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | PlotEllipse.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Localisation/PlotEllipse.m | 381 | utf_8 | 75ead0779e18e88a48b625e1d3e376f5 | %%%%%%%% Cavariance elipse ( Eig values) %%%%%%%%
function eH = PlotEllipse(x,P,nSigma)
eH = [];
P = P(1:2,1:2); % we are only interested in x and y part
x = x(1:2);
if(~any(diag(P)==0))
[V,D] = eig(P);
y = nSigma*[cos(0:0.1:2*pi);sin(0:0.1:2*pi)];
el = V*sqrtm(D)*y;
el = [el el(:,1)]+repmat(x,... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | DrawRobot.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Localisation/DrawRobot.m | 506 | utf_8 | a3ec6aa3dde44aa6ae5ae373e90d5cf4 | %-------- Drawing Vehicle -----%
function DrawRobot(Xr,col,ShiftTheta);
p=0.02; % percentage of axes size
a=axis;
l1=(a(2)-a(1))*p;
l2=(a(4)-a(3))*p;
P=[-1 1 0 -1; -1 -1 3 -1];%basic triangle
theta = Xr(3)-pi/2+ShiftTheta;%rotate to point along x axis (theta = 0)
c=cos(theta);
s=sin(theta);
P=[c -s; s c]*P... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | AngleWrap.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Mapping/AngleWrap.m | 173 | utf_8 | 6e4585850dc57bd5fbce3fd38a1a12da | %% Function to handle the transformation of angle
function angle = AngleWrap(angle)
if(angle>pi)
angle=angle-2*pi;
elseif(angle<-pi)
angle = angle+2*pi;
end; |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | tinv.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Mapping/tinv.m | 433 | utf_8 | bc8bf0fa4d4dc647e88ed009cc55657a | function tba=tinv(tab)
tba = zeros(size(tab));
for t=1:3:size(tab,1),
tba(t:t+2) = tinv1(tab(t:t+2));
end
end
%-------------------------------------------------------
function tba=tinv1(tab)
%
% calculates the inverse of one transformations
%-------------------------------------------------------
s = ... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | PlotEllipse.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Mapping/PlotEllipse.m | 373 | utf_8 | ae1d255d26224f21b23e8eb7ed6d6dc7 | %-------- Drawing Covariance -----%
function eH = PlotEllipse(x,P,nSigma,k)
eH = [];
P = P(1:2,1:2); % only plot x-y part
x = x(1:2);
if(~any(diag(P)==0))
[V,D] = eig(P);
y = nSigma*[cos(0:0.1:2*pi);sin(0:0.1:2*pi)];
el = V*sqrtm(D)*y;
el = [el el(:,1)]+repmat(x,1,size(el,2)+1);
eH = line(... |
github | Lumbrer/Feature-Based-SLAM-Basic-Tutorial-master | DrawRobot.m | .m | Feature-Based-SLAM-Basic-Tutorial-master/Mapping/DrawRobot.m | 506 | utf_8 | a3ec6aa3dde44aa6ae5ae373e90d5cf4 | %-------- Drawing Vehicle -----%
function DrawRobot(Xr,col,ShiftTheta);
p=0.02; % percentage of axes size
a=axis;
l1=(a(2)-a(1))*p;
l2=(a(4)-a(3))*p;
P=[-1 1 0 -1; -1 -1 3 -1];%basic triangle
theta = Xr(3)-pi/2+ShiftTheta;%rotate to point along x axis (theta = 0)
c=cos(theta);
s=sin(theta);
P=[c -s; s c]*P... |
github | DesignInformaticsLab/3D-CNN-master | rec_completion_test.m | .m | 3D-CNN-master/3DShapeNets/rec_completion_test.m | 9,746 | utf_8 | 07cacd80e8a61c09be571208f932b0cd | function [completed_data, predicted_label, energy] = rec_completion_test(model, test_data, mask, show, param)
% Given incomplete 3D shape(tsdf), recognition and completion test for
% multi-class model.
% Input the trained model, and some fixed masks, this function perform
% completion and recognition simultaneously. ... |
github | DesignInformaticsLab/3D-CNN-master | sample_test_extreme.m | .m | 3D-CNN-master/3DShapeNets/sample_test_extreme.m | 4,280 | utf_8 | bf8afa9a256fa45b4fc2a586d0a4d4fa | function [batch_data, batch_label] = sample_test_extreme(model, class)
% Gibbs sampling for multi-class models. Somewhat like
% sample_test_classification, but sampling process involves all layers
% in way that mimics the completion process(up down up down). If this
% sampling can give good results, completion performa... |
github | DesignInformaticsLab/3D-CNN-master | sample2TSDF_fast.m | .m | 3D-CNN-master/3DShapeNets/3D/sample2TSDF_fast.m | 4,833 | utf_8 | 1b8b173b6e256677d69cb771a60371a3 | function [new_TSDF] = sample2TSDF_fast(completed_samples, center, K, R, trans, halfWidth, R_cam, trans_cam, depth)
% Project completed samples to the surface specified by the new camera.
% completed_samples: input completions
% center: object center in world coordinate
% K: camera intrinsic
% R, trans: camera extrinsic... |
github | DesignInformaticsLab/3D-CNN-master | rgb_plane2rgb_world.m | .m | 3D-CNN-master/3DShapeNets/3D/rgb_plane2rgb_world.m | 793 | utf_8 | 72d4db5839a8ec7551642559d77d1205 | % Projects the depth points from the image plane to the 3D world
% coordinates.
%
% Args:
% imgDepth - depth map which has already been projected onto the RGB
% image plane, an HxW matrix where H and W are the height and
% width of the matrix, respectively.
%
% Returns:
% points3d - the po... |
github | DesignInformaticsLab/3D-CNN-master | off2im.m | .m | 3D-CNN-master/3DShapeNets/3D/off2im.m | 3,590 | utf_8 | 6a351ae3e5dc899523d0210c12dc492e | function [depth,K,crop] = off2im(offfile, ratio, xzRot, Rtilt, objx,objy, objz, modelsize,addfloor,enlargefloor)
% Render a depth map from a 3D mesh model provided by Shuran Song.
% calls RenderMex
% offfile: off filename
% ratio: set it 1
% xzRot: rotate angle of the 3D mesh model
% Rtilt: tilt angle of the 3D mesh m... |
github | DesignInformaticsLab/3D-CNN-master | get_aligned_point_cloud.m | .m | 3D-CNN-master/3DShapeNets/3D/get_aligned_point_cloud.m | 317 | utf_8 | 136c5c83adf6ef25381beb225f183526 | % Aligns the point cloud given a rotation matrix.
%
% Args:
% points3d - Nx3 point cloud.
% R - 3x3 rotation matrix.
%
% Returns:
% points3d - Nx3 aligned point cloud.
%
% Author: Nathan Silberman (silberman@cs.nyu.edu)
function points3d = get_aligned_point_cloud(points3d, R)
points3d = (R * points3d')';
end |
github | DesignInformaticsLab/3D-CNN-master | get_points_in_bb3d.m | .m | 3D-CNN-master/3DShapeNets/3D/get_points_in_bb3d.m | 1,183 | utf_8 | 94904b8d65614a7304ce14cf594507a0 | function [bbIdx, basis] = get_points_in_bb3d(points3d, bb3d)
% extract object point cloud from the 3d bounding box.
% modified from NYU code.
% points3d: point cloud from the scene.
% bb3d: object bounding box in 3d.
% each row is a orthogonal vector in basis matrix.
% column vector: x_world = basis' * x_bb + centroid... |
github | DesignInformaticsLab/3D-CNN-master | icp.m | .m | 3D-CNN-master/3DShapeNets/3D/icp.m | 18,342 | utf_8 | 283054154b9888ad4a7e24be82f8851c | function [TR, TT, ER, t] = icp(q,p,varargin)
% Perform the Iterative Closest Point algorithm on three dimensional point
% clouds.
%
% [TR, TT] = icp(q,p) returns the rotation matrix TR and translation
% vector TT that minimizes the distances from (TR * p + TT) to q.
% p is a 3xm matrix and q is a 3xn matrix.
%
% [TR,... |
github | DesignInformaticsLab/3D-CNN-master | TSDF.m | .m | 3D-CNN-master/3DShapeNets/3D/TSDF.m | 4,360 | utf_8 | c82e9998839007c787b441863cf88356 | function [all_gridDists, halfWidth] = TSDF(depth, K, center, R, trans, volume_size, pad_len, halfWidth, crop)
% This TSDF merges multiple depth map (and Rt)
% altogether into one TSDF. TSDF: 1 for object surface, 0 for empty spaces,
% -1 for unknown voxels.
% depth: depth map for the original view
% K: camera intrinsi... |
github | DesignInformaticsLab/3D-CNN-master | sigmoid.m | .m | 3D-CNN-master/3DShapeNets/my_code/sigmoid.m | 54 | utf_8 | fc270804a69bbdac0b1a492007a8a5ed |
function [y] = sigmoid(x)
y = 1 ./ (1 + exp(-x));
end |
github | DesignInformaticsLab/3D-CNN-master | filter_visual.m | .m | 3D-CNN-master/3DShapeNets/my_code/filter_visual.m | 2,491 | utf_8 | 5a1b824e2cdd2784ac99cff18842b2e9 | % function filter_visual()
% load('./pretrained_model.mat');
% w = rec_conv(model,3); % filter of which rbm
% save('w_projected.mat','w');
% % plot_filter(w)
% for i=1:10:16
% figure;
% cnt = 1;
% for t=6:-1:1
% subplot(2,3,cnt);
% show_sample(squeeze(w(i,:,:,:)),t)
% title(strcat('t... |
github | DesignInformaticsLab/3D-CNN-master | reconstruction.m | .m | 3D-CNN-master/3DShapeNets/my_code/reconstruction.m | 3,983 | utf_8 | 8bd9a1fdbb59fc1822ebdaa1d7ec613b | % use construction test the accuracy of RBM
function reconstruction()
kernels
% reset(gpuDevice(1));
run('setup_paths.m')
model = load('pretrained_model.mat');
filename2 = './my_code/more/pot_train_sal.mat';
% filename2 = './volumetric_data/chair/30/train/chair_000000182_8.mat';
% filename2 = './volumetric_data/my_cup_... |
github | DesignInformaticsLab/3D-CNN-master | write_input_data.m | .m | 3D-CNN-master/3DShapeNets/util/write_input_data.m | 2,938 | utf_8 | 0c64c3e1f8e32d06265c455d40d047c8 | function write_input_data(off_path, data_path, classes, volume_size, pad_size, angle_inc)
% Put the mesh object in a volume grid and save the volumetric
% represenation file.
% This is the input volumetric data for 3D ShapeNets.
% off_path: root off data folder
% data_path: destination volumetric data folder
phases = ... |
github | DesignInformaticsLab/3D-CNN-master | vol3d.m | .m | 3D-CNN-master/3DShapeNets/util/vol3d.m | 7,609 | utf_8 | 5746b9360eca710bdf9c8cbe6053c65a | function [model] = vol3d(varargin)
%H = VOL3D Volume render 3-D data.
% VOL3D uses the orthogonal plane 2-D texture mapping technique for
% volume rending 3-D data in OpenGL. Use the 'texture' option to fine
% tune the texture mapping technique. This function is best used with
% fast OpenGL hardware.
%
% vol3... |
github | DesignInformaticsLab/3D-CNN-master | parse_json.m | .m | 3D-CNN-master/3DShapeNets/util/parse_json.m | 5,774 | utf_8 | c1a6d0a6fd907de1891f9aca3201fbad | function [data json] = parse_json(json)
% [DATA JSON] = PARSE_JSON(json)
% This function parses a JSON string and returns a cell array with the
% parsed data. JSON objects are converted to structures and JSON arrays are
% converted to cell arrays.
%
% Example:
% google_search = 'http://ajax.googleapis.com/ajax/s... |
github | DesignInformaticsLab/3D-CNN-master | wake_sleep_CD.m | .m | 3D-CNN-master/3DShapeNets/generative/wake_sleep_CD.m | 12,626 | utf_8 | 799877402e4b919d6b501e2f128edf27 | function [model] = wake_sleep_CD(model, data_list, param)
% Generatively fine-tuning the model using the wake sleep algorithm.
% This is the version that I changed, tuning the weight without untying the
% weights. The top RBM keeps a negative persistant chain.
% data_list: an array of filenames returned by balance_data... |
github | DesignInformaticsLab/3D-CNN-master | crbm2.m | .m | 3D-CNN-master/3DShapeNets/generative/crbm2.m | 6,610 | utf_8 | e07ed01d095942f034c50cd33eb594b4 | function [model] = crbm2(model, data_list, param)
% Convolutional RBM training of the second layer.
% The CRBM of second layer uses different convolution CUDA kernels with the
% third and fourth layer.
% data_list: contains an array of filenames. Returned by balance_data.m
% param: training parameters set in run_p... |
github | DesignInformaticsLab/3D-CNN-master | get_cross_entropy_all.m | .m | 3D-CNN-master/3DShapeNets/generative/get_cross_entropy_all.m | 6,937 | utf_8 | 2717400c32365394940c444ce3155ddb | function err = get_cross_entropy_all(model, new_list, label, to_layer)
% Reconstruction cost of the whole model from input to the layer to_layer.
% Propagate the data to the layer (to_layer) and propagate it down.
% Examine Its reconstruction error.
global kConv_backward kConv_backward_c kConv_forward2 kConv_forward_... |
github | DesignInformaticsLab/3D-CNN-master | crbm.m | .m | 3D-CNN-master/3DShapeNets/generative/crbm.m | 5,635 | utf_8 | 616cd82f21a75512a97343d00533a578 | function [model] = crbm(model, data_list, param)
% Convolutional RBM training of the third and fourth layer.
% data_list: contains an array of filenames. Returned by balance_data.m
% param: training parameters set in run_pretrain.m
global kConv_backward_c kConv_forward_c kConv_weight_c;
lr = param.lr;
l = param.layer... |
github | DesignInformaticsLab/3D-CNN-master | rbm_last.m | .m | 3D-CNN-master/3DShapeNets/generative/rbm_last.m | 6,822 | utf_8 | 39019f8d9e650862e20eedfec21bc8cb | function [model, hidden_prob] = rbm_last(model, data_list, data, param)
% FPCD RBM training for the joint training of data + label(last top layer).
l = param.layer;
lr = param.lr;
batch_size = param.batch_size;
wd = param.weight_decay;
persistant = param.persistant;
sparse_damping = param.sparse_damping;
spar... |
github | DesignInformaticsLab/3D-CNN-master | get_cross_entropy.m | .m | 3D-CNN-master/3DShapeNets/generative/get_cross_entropy.m | 4,847 | utf_8 | aae5fb9ce90d01f708d5af8a500cd7c1 | function err = get_cross_entropy(model, data, l)
% cross-entropy(reconstruction error, I use L2) of each layer. This is used
% to monitor the pretraining progress.
% data: can be real data(for rbm) or data filelist(crbm).
global kConv_backward kConv_backward_c kConv_forward2 kConv_forward_c;
fraction = 5;
if l == 2
... |
github | DesignInformaticsLab/3D-CNN-master | NBV_onestep.m | .m | 3D-CNN-master/3DShapeNets/generative/NBV_onestep.m | 7,224 | utf_8 | ed3dac19f54639cba26c9dfcb3d9c7bd | function [prediction_o, prediction_next, NBV_v, RC_v, RAND_v, FUR_v]= NBV_onestep(model, filename, angle_inc)
% Next-Best-View for one step. Given a depth map, calculate the TSDF to
% reconstruct the 3D partial surface and other unknown spaces. Then decide
% the NBV with the least uncertainty from a set of possible cam... |
github | DesignInformaticsLab/3D-CNN-master | wake_sleep.m | .m | 3D-CNN-master/3DShapeNets/generative/wake_sleep.m | 14,456 | utf_8 | 1a5422100506a24de2913b0dfdb76e6f | function [model]= wake_sleep(model, data_list, param)
% Generatively fine-tuning the model using the wake sleep algorithm.
% This is the original wake_sleep algorithm, untying the generative weights
% and recognition weights
% data_list: an array of filenames returned by balance_data.m
% param: fine-tuning parameters s... |
github | austrin/max-bisection-analysis-master | trho.m | .m | max-bisection-analysis-master/heuristics/trho.m | 530 | utf_8 | 53de7a3cfff5535ee45d27c2ed9c45ff | % trho(mu1, mu2, rho) = (rho-mu1*mu2)/sqrt((1-mu1^2)(1-mu2^2))
%
% The arguments can be matrices, in which case they must all be of the
% same dimension.
%
% Requires:
% -1 <= mu1 <= 1
% -1 <= mu2 <= 1
% -1 <= rho <= 1
% -mu1-mu2-rho <= 1
% -mu1+mu2+rho <= 1
% mu1+mu2-rho <= 1
% mu1-mu2+rho <= 1
%
funct... |
github | austrin/max-bisection-analysis-master | alpha_cf.m | .m | max-bisection-analysis-master/heuristics/alpha_cf.m | 1,608 | utf_8 | f49644db530f3e352210bce4b31095f2 | % Compute ratio alpha_{c,f}(mu1, mu2, rho) from the pairing algorithm.
%
% mu1, mu2, rho form a configuration
% 0 <= c <= 1 is the linear scaling factor
% f: [0,1] -> [0,1] is the "bias boosting" function
% f must be able to take matrices as inputs.
%
% mu1, mu2, rho may be matrices, in which case they must all hav... |
github | austrin/max-bisection-analysis-master | alpha.m | .m | max-bisection-analysis-master/heuristics/alpha.m | 676 | utf_8 | dd07dae41e18a91c709069fa1ba277ae | % alpha(mu1, mu2, rho, r1, r2) = (2(1-Lambda(r1, r2, trho)))/(1-rho)
% where trho = (rho-mu1*mu2)/sqrt((1-mu1^2)(1-mu2^2))
%
% Evaluates a (set of) roundings on a (set of) configurations.
%
% The arguments can be matrices, in which case they must all be of the
% same dimension.
%
% Requires:
% -1 <= mu1 <= 1
% -1 <... |
github | austrin/max-bisection-analysis-master | bvnl.m | .m | max-bisection-analysis-master/heuristics/bvnl.m | 4,611 | utf_8 | 8ea13d0b904a12cc1c407ed0bb299b9d | function p = bvnl( dh, dk, r )
%
% a function for computing bivariate normal probabilities.
% bvnl calculates the probability that x < dh and y < dk.
% parameters
% dh 1st upper integration limit
% dk 2nd upper integration limit
% r correlation coefficient
%
% this function is based on ... |
github | austrin/max-bisection-analysis-master | alpha_pairing.m | .m | max-bisection-analysis-master/heuristics/alpha_pairing.m | 2,205 | utf_8 | 6398dbcc1c5574831f020ce1ecfa4e72 | % Compute ratio alpha(c, f) from the pairing algorithm.
%
% Searches for worst configuration by using Matlab's Optimization
% Toolbox, decrease risk of error by running the search many times
% with different starting points.
%
% Three optional arguments:
% - trials (default 50): number of searches to run.
% - eps (defa... |
github | austrin/max-bisection-analysis-master | Gamma.m | .m | max-bisection-analysis-master/heuristics/Gamma.m | 729 | utf_8 | 0e32136b5c0b6fdbcc4939a5d67eb2c4 | % Gamma(x, y, rho) = Pr[X <= Phi^{-1}(x) and Y <= Phi^{-1}(y)]
% where (X,Y) are jointly normal random variables with mean 0, variance 1, and covariance rho
%
% The arguments can be matrices, in which case they must all be of the
% same dimension.
%
% Requires:
% 0 <= x <= 1
% 0 <= y <= 1
% -1 <= rho <= 1
%
functi... |
github | austrin/max-bisection-analysis-master | normalinv.m | .m | max-bisection-analysis-master/heuristics/normalinv.m | 134 | utf_8 | f703309ffc0ee703230b7197f02af977 | % Phi^{-1}(x), where Phi(x) is the cdf of a standard N(0,1) Gaussian
function phiinv=normalinv(x)
phiinv = sqrt(2)*erfinv(2*x-1);
|
github | austrin/max-bisection-analysis-master | gendata_pairing_contour.m | .m | max-bisection-analysis-master/heuristics/gendata_pairing_contour.m | 2,026 | utf_8 | 4da30c64ce2d6d3c0ba85454c9649698 | % Computes the ratio alpha_{c,f}(mu1, mu2, rho) from the pairing algorithm along the
% boundary of the polytope
%
% Parameters:
% 0 <= c <= 1 is the linear scaling factor
% f: [0,1] -> [0,1] is the "bias boosting" function
% f must be able to take matrices as inputs.
% n integer giving the granularity (step size w... |
github | austrin/max-bisection-analysis-master | Lambda.m | .m | max-bisection-analysis-master/heuristics/Lambda.m | 315 | utf_8 | 91ac8728cb0bc69d1b69984b478231b3 | % Lambda(r1, r2, trho) = 2Gamma((1-r1)/2, (1-r2)/2, trho) + (r1+r2)/2
%
% The arguments can be matrices, in which case they must all be of the
% same dimension.
%
% Requires:
% -1 <= r1 <= 1
% -1 <= r2 <= 1
% -1 <= trho <= 1
function L = Lambda(r1, r2, trho)
L = 2*Gamma((1-r1)/2, (1-r2)/2, trho) + (r1+r2)/2;
|
github | austrin/max-bisection-analysis-master | alpha_I.m | .m | max-bisection-analysis-master/heuristics/alpha_I.m | 1,691 | utf_8 | 2740e96c7e52f14caac75f109cb9dd49 | % Compute ratio alpha(mu1, mu2, rho, I1, I2) from the pairing algorithm.
%
% mu1, mu2, rho form a configuration
% I1, I2 are intervals (in the form of 2x1 or 1x2-dimensional vectors)
%
% Note: this function does NOT accept matrix arguments.
%
function ratio = alpha_I(mu1, mu2, rho, I1, I2)
tr = trho(mu1, mu2, rho);
... |
github | austrin/max-bisection-analysis-master | Phi.m | .m | max-bisection-analysis-master/heuristics/Phi.m | 98 | utf_8 | fb52b97c2052a31efbabbd44bdd689d5 | % Phi(x), the cdf of a standard N(0,1) Gaussian
function y = Phi(x)
y = (1 + erf(x/sqrt(2)))/2; |
github | austrin/max-bisection-analysis-master | alpha_linear.m | .m | max-bisection-analysis-master/heuristics/alpha_linear.m | 2,126 | utf_8 | 8e96d0e5ac0e39b6c7c911fb3bdb85cc | % Compute ratio alpha(c) from the algorithm using linear biases.
%
% Searches for worst configuration by using Matlab's Optimization
% Toolbox, decrease risk of error by running the search many times
% with different starting points.
%
% Three optional arguments:
% - trials (default 50): number of searches to run.
% - ... |
github | 123chengbo/ssd-windows-master | classification_demo.m | .m | ssd-windows-master/matlab/demo/classification_demo.m | 5,412 | utf_8 | 8f46deabe6cde287c4759f3bc8b7f819 | function [scores, maxlabel] = classification_demo(im, use_gpu)
% [scores, maxlabel] = classification_demo(im, use_gpu)
%
% Image classification demo using BVLC CaffeNet.
%
% IMPORTANT: before you run this demo, you should download BVLC CaffeNet
% from Model Zoo (http://caffe.berkeleyvision.org/model_zoo.html)
%
% *****... |
github | geodesymiami/GeodMod-master | add_shade2Data.m | .m | GeodMod-master/PlotDatalib/add_shade2Data.m | 2,761 | utf_8 | 1285fbf46e56aef54512d2ce365df6bb |
function [igram]=add_shade2Data(Igram,ShadeFile,lopt)
%Prepare dem and save it in the igram structure
%
% Igram = igram structure
% ShadeFile = DEM (.jpg only)
%
% Noel Gourmelen October 2005
% Falk Amelung 15 November
% Now allows lopt
% putting shade int... |
github | geodesymiami/GeodMod-master | maxmin.m | .m | GeodMod-master/PlotDatalib/maxmin.m | 157 | utf_8 | 255c63cbb96f3e23b89c3ab4a58267c6 | %Max value of array p1:
function [Maxf,minf]= maxmin (file);
maxp1=max(file); %by columns
Maxf=max(maxp1);
minp1=min(file); %by columns
minf=min(minp1); |
github | geodesymiami/GeodMod-master | plot_NaNbackground.m | .m | GeodMod-master/PlotDatalib/plot_NaNbackground.m | 1,119 | utf_8 | 3ce3459b137e14f6433277799743666b |
function plot_NaNbackground(data,opt)
%
% Plot data containg NaNs. Only data will be used to scale the colorscale
%
% data : data!
%
% opt:
%
% cmap : Colormap
% background_color : Example -> [0] for white (default)
%
%
% Noel Gourmelen - March 2009
%
defaultopt=struct( ... |
github | geodesymiami/GeodMod-master | beachball.m | .m | GeodMod-master/PlotDatalib/beachball.m | 8,414 | utf_8 | 5011d1c1ca3f37b8393c7c29f59146d2 | function handle = beachball(strike,dip,rake,x0,y0,radius,color,handle)
% Usage: handle = beachball(strike,dip,rake,x0,y0,radius,color,handle)
%
% Plot a lower-hemisphere focal mechanism centered at x0,y0
% with radius radius.
% handle is an optional argument. If specified, and if handle
% points to an existing be... |
github | geodesymiami/GeodMod-master | anneal_parfortry.m | .m | GeodMod-master/inver/anneal_parfortry.m | 8,042 | utf_8 | 240767d298ea2ccd0c03678e3f12c089 | function [mhat,F,model,energy,count]=anneal(FUN,bounds,OPTIONS,varargin)
% anneal - simulated annealing by Peter Cervelli
%ANNEAL [mhat,F,model,energy,count]=anneal(FUN,bounds,OPTIONS,x1,x2...,xn)
%
%Simulated annealing algorithm that tries to find a minimum to the function 'FUN'.
%
%INPUTS:
%
%'FU... |
github | geodesymiami/GeodMod-master | plot_gridsearch.m | .m | GeodMod-master/inver/plot_gridsearch.m | 5,771 | utf_8 | 787f96937f9a52c0ef2a0dbe61a8700a | function [momin,ppd1d,ppd2d]=plot_gridsearch(models,energy,bounds,gibbsopt,inverseopt,objfuncopt,momin,ppd1d,ppd2d)
%
% PLOT_GIBBS plots 1-D and 2-D marginal probability distributions
%
% usage: [momin,ppd1d,ppd2d]=plot_gibbs(models,energy,bounds,gibbsopt,momin,ppd1d,ppd2d)
%
% PLOT_GIBBS(MODELS,ENERGY,BOUNDS,OP... |
github | geodesymiami/GeodMod-master | anneal.m | .m | GeodMod-master/inver/anneal.m | 8,321 | utf_8 | f2dc367f2eaa118977033f119180ff3a | function [mhat,F,model,energy,count]=anneal(FUN,bounds,OPTIONS,varargin)
% anneal - simulated annealing by Peter Cervelli
%ANNEAL [mhat,F,model,energy,count]=anneal(FUN,bounds,OPTIONS,x1,x2...,xn)
%
%Simulated annealing algorithm that tries to find a minimum to the function 'FUN'.
%
%INPUTS:
%
%'F... |
github | geodesymiami/GeodMod-master | pg.m | .m | GeodMod-master/inver/gibbslib/pg.m | 7,841 | utf_8 | ebb087b38be71a278c2ccf24e0389e17 | function [momin,ppd1d,ppd2d]=plot_gibbs(models,energy,bounds,gibbsopt,momin,ppd1d,ppd2d)
%
% PLOT_GIBBS plots 1-D and 2-D marginal probability distributions
%
% usage: [momin,ppd1d,ppd2d]=plot_gibbs(models,energy,bounds,gibbsopt,momin,ppd1d,ppd2d)
%
% PLOT_GIBBS(MODELS,ENERGY,BOUNDS,OPT) ca... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.