run_tracker_demo.m

%   Robust Visual Tracking via Hierarchical Convolutional Features
%   Chao Ma, Jia-Bin Huang, Xiaokang Yang, and Ming-Hsuan Yang
%   IEEE Transactions on Pattern Analysis and Machine Intellegince


function [precision, fps] = run_tracker_demo(video, show_visualization, show_plots)

% run_HCFTstar_demo:
% process a sequence using HCFTstar (Correlation filter tracking with convolutional features)
%
% Input:
%     - video:        sequence name
%     - show_visualization:   whether to show results
%     - show_plots: whether to show tracking precision
% Output:
%     - precision: tracking results with bounding boxes over time
%     - fps: tracking results with FPS
%
%   It is provided for educational/researrch purpose only.
%   If you find the software useful, please consider cite our paper.



dbstop if error;

%path to the OTB videos (you'll be able to choose one with the GUI).
base_path ='./OTB_dataset';

close all;

addpath('utility','train');

% Path to MatConvNet. Please run external/matconvnet/vl_compilenn.m to
% set up the MatConvNet

addpath './matconvnet/matlab'
vl_setupnn();

% Where the 'imagenet-vgg-verydeep-19.mat' file is
addpath './vgg_model'

addpath(genpath('edgesbox'));
addpath(genpath('piotr_toolbox'));
addpath(genpath('Diagnose'))

% Default settings
if nargin < 1, video = 'choose'; end
if nargin < 2, show_visualization = ~strcmp(video, 'all'); end
if nargin < 3, show_plots = ~strcmp(video, 'all'); end

% Extra area surrounding the target
padding = struct('generic', 1.8, 'large', 1, 'height', 0.4);

lambda = 1e-4;              % Regularization parameter (see Eqn 3 in our paper)
output_sigma_factor = 0.1;  % Spatial bandwidth (proportional to the target size)

interp_factor = 0.01;       % Model learning rate (see Eqn 6a, 6b)
cell_size = 4;              % Spatial cell size
config.kernel_sigma = 1;
config.motion_thresh= 0.181; %0.25 for singer2 0.32;%0.15
config.appearance_thresh=0.38; %0.38
config.features.hog_orientations = 9;
config.features.cell_size = 4;   % size of hog grid cell		
config.features.window_size = 6; % size of local region for intensity historgram  
config.features.nbins=8; 
global enableGPU;
enableGPU = true;

switch video
    case 'choose'
        % Ask the user for selecting the video, then call self with that video name.
      %  matlabpool open local 8;
        video = choose_video(base_path);
        if ~isempty(video)
            % Start tracking
            [precision, fps] = run_tracker_demo(video, show_visualization, show_plots);
            
            if nargout == 0  % Don't output precision as an argument
                clear precision
            end
        end
        
    case 'all'
        %all videos, call self with each video name.
        
        %only keep valid directory names
        dirs = dir(base_path);        videos = {dirs.name};
        videos(strcmp('.', videos) | strcmp('..', videos) | ...
            strcmp('anno', videos) | ~[dirs.isdir]) = [];
        
        % Note: the 'Jogging' sequence has 2 targets, create one entry for each.
        % we could make this more general if multiple targets './top-down/'per video
        % becomes a common occurence.
        
        %=========================================================================
        % Uncomment following scripts if you test on the entire bechmark
%                 videos(strcmpi('Jogging', videos)) = [];
%                 videos(end+1:end+2) = {'Jogging.1', 'Jogging.2'};
        %
%                 videos(strcmpi('Skating2', videos))=[];
%                 videos(end+1:end+2)={'Skating2.1', 'Skating2.2'};
        %=========================================================================
        
        all_precisions = zeros(numel(videos),1);  % to compute averages
        all_fps = zeros(numel(videos),1);
        
        %poolobj = gcp;
        poolobj=gcp('nocreate');
        parfor k = 1:numel(videos)
            %if exist([result_path videos{k} '.mat'],'file'), continue; end
            [prec, all_fps(k)] = run_tracker_demo(videos{k}, show_visualization, show_plots);
            
            all_precisions(k) = prec;
        end
        
        delete(poolobj);
        
        %compute average precision at 20px, and FPS
        mean_precision = mean(all_precisions);
        
        fps = mean(all_fps);
        
        fprintf('\nAverage precision (20px):% 1.3f, Average FPS:% 4.2f\n\n', mean_precision, fps)
        save([result_path  'average' '.mat'],'mean_precision');
        if nargout > 0
            precision = mean_precision;
        end
        
    otherwise
        % We were given the name of a single video to process.
        % get image file names, initial state, and ground truth for evaluation
        [img_files, pos, target_sz, ground_truth, video_path] = load_video_info(base_path, video);
        
        % Call tracker function with all the relevant parameters
%         [positions, time] = tracker_ensemble(video_path, img_files, pos, target_sz, ...
%             padding, lambda, output_sigma_factor, interp_factor, ...
%             cell_size, show_visualization);
        [positions, time,rect_position] = tracker_HCFTstar(video_path, img_files, pos, target_sz, ...
            padding, lambda, output_sigma_factor, interp_factor, ...
            cell_size, show_visualization,config); %tracker_ensemble_RPnew1 
        
        % Calculate and show precision plot, as well as frames-per-second
        precisions = precision_plot(positions, ground_truth, video, show_plots);
        fps = numel(img_files) / time;
        results.type = 'rect'; 
        results.res = rect_position;%each row is a rectangle 
        results.len = size(precisions,1);
        results.fps = fps; 

        fprintf('%12s - Precision (20px):% 1.3f, FPS:% 4.2f\n', video, precisions(20), fps)
        precision=precisions(20);  

        if nargout > 0
            %return precisions at a 20 pixels threshold
            precision = precisions(20);
        end
end
end