add recognition

recognition/.gitignore

@@ -0,0 +1,107 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+env/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# dotenv
+.env
+
+# virtualenv
+.venv
+venv/
+ENV/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+WARD1.0
+*.asv
+data.mat
+results.mat
+performances.mat
+*.mat

recognition/activity_recognition.m

@@ -0,0 +1,57 @@
+function results = activity_recognition(lambda, sigma)
+    close all
+    load('data.mat');
+
+    if nargin < 2
+        lambda = 0.0015;
+        sigma = 1;
+    end
+
+    sequence = 76:125;
+    time_kcc = zeros(13,1);
+    time_dtw = zeros(13,1);
+
+    %% examples
+    for subject = 1:size(data,1)
+        for activity = 1:size(data, 2)
+            for samples = 1:5%size(data, 3) %only one subject has the 6th sample in only one activity
+                if isempty(data{subject, activity, samples})
+                    continue;
+                end
+
+                train = data{subject, activity, samples}(sequence, :);
+                tic
+                correlator = kcc_train(train, lambda, sigma);
+                time_kcc(activity) = time_kcc(activity) + toc;
+
+                for activity_test = 1:size(data, 2)
+                    for samples_test = 1:5%size(data, 3) %only one subject has the 6th sample in only one activity
+                        if isempty(data{subject, activity_test, samples_test})
+                            continue;
+                        end
+
+                        test = data{subject, activity_test, samples_test}(sequence, :);
+
+                        tic
+                        response(subject, 5*(activity-1)+ samples, 5*(activity_test-1)+ samples_test) = kcc(test, correlator);
+                        time_kcc(activity_test) = time_kcc(activity_test) + toc;
+
+                        tic
+%                         distance(subject, 5*(activity-1)+ samples, 5*(activity_test-1)+ samples_test) = 0;
+                        distance(subject, 5*(activity-1)+ samples, 5*(activity_test-1)+ samples_test) = dtw(train', test');
+                        time_dtw(activity_test) = time_dtw(activity_test) + toc;
+                    end
+                end
+            end    
+        end
+    end
+
+    time_use = [time_kcc, time_dtw];
+    filename = 'results.mat';
+    save(filename, 'response', 'distance', 'time_use');
+    [accuracy_kcc, accuracy_dtw] = show_results(filename);
+    results = [accuracy_dtw, accuracy_kcc, time_dtw, time_kcc, sigma, lambda];
+    fprintf('accuracy_dtw: %f; accuracy_kcc: %f; time_dtw: %f, time_kcc: %f; sigma: %f lambda: %f\n', ...
+        results(1),  results(2), results(3), results(4), results(5), results(6));
+
+end

recognition/analyse_parameters.m

@@ -0,0 +1,6 @@
+% clc
+% clear
+load('performances_refine_refin.mat');
+plot3(performances(:,end-1), performances(:,end), performances(:,2),'.')
+performances(:,[end-1,end,2])
+max(performances(:,2))

recognition/comparison_results.txt

@@ -0,0 +1,17 @@
+\begin{tabular}{|l|l|l|l|}
+\hline
+74&63.5&119.0911&161.0318\\\hline
+75.5&66&116.525&156.9842\\\hline
+72&72.5&116.66&157.161\\\hline
+96&87&116.6862&157.8734\\\hline
+93&80&116.7369&157.3694\\\hline
+94.5&84&113.9792&153.9453\\\hline
+95.5&80&115.4979&155.2064\\\hline
+95&83&116.6258&157.0647\\\hline
+94.5&80&116.5067&157.0048\\\hline
+99.5&79&116.4776&157.2347\\\hline
+100&95&116.5865&157.0245\\\hline
+97&86&116.4139&157.0018\\\hline
+92.5&69&115.3224&155.4016\\\hline
+90.6923&78.8462&116.393&156.9464\\\hline
+\end{tabular}

recognition/data_conversion.m

@@ -0,0 +1,41 @@
+clear; clc
+% You need to download the WARD dataset to current folder.
+% 
+folder = './WARD1.0/';
+addpath('./natsortfiles/')
+files = dir(folder);
+fileIndex = find([files.isdir]);
+fileIndex = fileIndex(3:end); 
+files = files(fileIndex);
+data={};
+
+files = natsortfiles({files(:).name});
+
+for i = 1:length(files)
+    subfolder = files(i); 
+    matfolders = dir(strcat(folder,subfolder{1}));
+    matIndex = find(~[matfolders.isdir]);
+    matfolders = matfolders(matIndex);
+    matfolder_files = natsortfiles({matfolders(:).name});
+
+    for j = 1:length(matfolder_files) 
+        matfiles = matfolder_files(j);
+        filename = strcat(folder, subfolder,'/',matfiles{1});
+        wd = load(filename{1});
+        reading=[];
+        for k=1:5
+            col = wd.WearableData.Reading{k};
+            col(isinf (col))=0;
+            col(isnan (col))=0;
+            tran = col-mean(col);
+            reading = [reading, tran./2./max(abs(tran))+0.5];
+            reading(isnan(reading))=0;
+        end
+        number = sscanf(filename{1},'./WARD1.0/Subject%d/a%dt%d.mat');
+        data{number(1),number(2), number(3)} = reading;
+%         data{i,j}.name = filename; 
+    end
+end 
+
+save('data.mat', 'data');
+clear

recognition/find_parameter.m

@@ -0,0 +1,12 @@
+clc
+clear
+lambda = 0.0014:0.0001:0.0020;
+sigma =  0.5:0.1:1.1;
+
+performances=[];
+for i = 1:numel(lambda)
+    for j = 1:numel(sigma)
+        performance = activity_recognition(lambda(i), sigma(j));
+        performances = [performance; performances];
+    end
+end

recognition/gaussian_kernel.m

@@ -0,0 +1,11 @@
+function kf = gaussian_kernel(xf, yf, sigma)
+	% Calculating the gaussian kernel vertor
+    % Copyright Wang Chen, Nanyang Technoglogical University
+
+	N = numel(xf);
+	xx = xf(:)' * xf(:) / N;
+	yy = yf(:)' * yf(:) / N;
+	xy = mean((ifft(xf .* conj(yf))),2);
+	kf = exp(-1 / sigma^2 * (xx + yy - 2 * xy) / N);
+end
+

recognition/kcc.m

@@ -0,0 +1,9 @@
+function [response, output] = kcc(test, correlator)
+	% Calculating the response
+    % Copyright Wang Chen, Nanyang Technoglogical University
+    test_fft = fft(test);
+    kernel_fft = fft(gaussian_kernel(test_fft, correlator.sample_fft, correlator.sigma));
+    output = abs(ifft(correlator.correlator_fft.*kernel_fft));
+    response = max(output);%/sum(output);
+end
+

recognition/kcc_train.m

@@ -0,0 +1,17 @@
+function correlator = kcc_train(sample, lambda, sigma)
+	% Training the correlator based on one sample
+    % Copyright Wang Chen, Nanyang Technoglogical University
+    if nargin < 3
+        sigma = 0.3;
+    end
+    if nargin < 2
+        lambda = 0.1;
+    end
+
+    correlator.sigma = sigma;
+    target_fft = ones(size(sample,1),1);
+    correlator.sample_fft = fft(sample);
+    kernel_fft = fft(gaussian_kernel(correlator.sample_fft, correlator.sample_fft, correlator.sigma));
+    correlator.correlator_fft = target_fft./(kernel_fft + lambda);
+end
+

recognition/matrix2latex/license.txt

@@ -0,0 +1,24 @@
+Copyright (c) 2009, Moritz Koehler
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+    * Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+    * Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the distribution
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGE.