main_video_size.m

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%% main function for collectiveness %%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% May 15, 2014, Jing Shao
% If you use this code, please cite the paper:
% J. Shao, C. C. Loy, X. Wang, "Scene-Independent Group Profiling in Crowd", CVPR, 2014.
clc;clear;close all

%% Descriptor -- Group size 
path = '.\';
path_gr = [path,'result_groupDet_new\'];
path_xls = [path, 'video_info_t0.xls'];
[~,~,xls] = xlsread(path_xls);
path_img_dir = xls(2:end,1);

% initialization and parameter setting
group_size_th = 25; 
hist_bin = [0:0.2:1];

gr_size = [];
for file_n = 1 : length(path_img_dir)
    file_name = path_img_dir{file_n};
    fprintf('Group descriptor "GroupSize" for [%d:%s].\n', file_n, file_name);
    
    %% load collective result from group detection
    load([path_gr, '\trkClusterTimeLine_1_', file_name, '.mat'], 'trkClusterTimeLine');
    load([path_gr, '\trks_', file_name, '.mat'], 'trks');
    load([path_gr, '\A_1_', file_name, '.mat'], 'A');
    load([path_gr, '\color_1_', file_name, '.mat'], 'color_ind')
    
    %%
    trkClusterNumTime = max(trkClusterTimeLine);
    [trkTime, lenTime, nTrks, trkTimeLine] = fun_trkInfo(trks);
    t_seq = find(trkClusterNumTime ~= 0);
    
    %% Do not need too long time (can be tuned)
    loca = cellfun(@findstr, xls(:,1), repmat({file_name}, size(xls(:,1))), 'UniformOutput', false);
    [t_loc, ~, ~] = find(~cellfun(@isempty, loca) == 1);
    t_start = fun_cell2num(xls(t_loc,5));
    t_end = min(t_seq(end),fun_cell2num(xls(t_loc,6)));
    
    %% groupSize computation
    gr_size = [];
    for curTime = t_start : t_end
        t_oo = find(t_seq==curTime);
        if isempty(t_oo)
            continue;
        end
        cur_color_ind = color_ind{t_oo};
        
        % prepare data
        cur_trk_ind = find(trkClusterTimeLine(:,curTime)~=0);
        cur_gr_ind = trkClusterTimeLine(cur_trk_ind,curTime);
        data = fun_curX(trks, nTrks, trkTime, curTime, cur_trk_ind);
        
        % preprocess data
        [cur_trk_ind, cur_gr_ind, data] = fun_curX_preprocess(data, cur_gr_ind, cur_trk_ind);
        clusterValue = unique(cur_gr_ind);
        
        for grSele = 1 : length(cur_color_ind)
            clusterV = cur_color_ind(grSele);
            ind = find(cur_gr_ind == clusterV);
            subdata = data(ind,:);
            gr_size{clusterV,curTime} = size(subdata,1);
        end
    end
    
    %% record: only groupSize larger than threshold
    group_size_mean = (sum(cellfun(@sum, gr_size),2))./(sum(~cellfun(@isempty,gr_size),2));
    group_size_max = []; group_size_sum = []; group_size_mean_all = [];
    for gr_n = 1 :length(group_size_mean)
        if group_size_mean(gr_n) > group_size_th
            group_size_max = [group_size_max; group_size_mean(gr_n)./max(group_size_mean)];
            group_size_sum = [group_size_sum; group_size_mean(gr_n)./sum(group_size_mean)];
        end
    end
    group_size(file_n).gr_size_max_v = hist(group_size_max, hist_bin);
    group_size(file_n).gr_size_sum_v = hist(group_size_sum, hist_bin);
    % other records
    group_size(file_n).gr_size_all = group_size_max.*max(group_size_mean);
    group_size(file_n).file_name = file_name;
    
    fprintf('Done!\n');
end

path_result = [path, 'result_groupDescr_new\'];
mkdir(path_result);
save([path_result,'groupSize.mat'], 'group_size')