1.算法概述
1.正确版本组合：Win7+Matlab R2015b+CUDA7.5+vs2013
CUDA7.5下载地址为：
developer.download.nvidia.com/compute/cud…
vs2013要专业版。
如下所示：
全部安装好之后，做如下操作：
2. CPU配置，运行CNN工具箱中的
然后再运行
可以完成CPP文件的编译。
3.编译成功后，会产生
这些必须在电脑上编译，否则用别人复制的，如果配置不一样，可能会报错。
4.GPU配置：
安装cudnn：developer.nvidia.com/rdp/cudnn-a…
然后mex -setup下，操作和CPU一样。
然后执行matlab程序：
vl_setupnn;
vl_compilenn('enableGpu', true,'cudaRoot', 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v7.5','cudaMethod', 'nvcc', 'enableCudnn', 'true','cudnnRoot', 'E:\A_2016_FPGA_Test\FC_tracking\A_FC\matconvnet-1.0-beta20\local\cudnn');
红色部分为路径
5.然后运行程序，就可以自动运行了，原来程序的运行非常麻烦，定义了这个程序，可以一步运行处结果。
2.仿真效果预览
Win7+Matlab R2015b+CUDA7.5+vs2013
运行后，如果可以出现如下结果：
3.核心MATLAB代码预览
`function bboxes = tracker(varargin)
%TRACKER
% is the main function that performs the tracking loop
% Default parameters are overwritten by VARARGIN
%
% Luca Bertinetto, Jack Valmadre, Joao F. Henriques, 2016
% -------------------------------------------------------------------------------------------------
% These are the default hyper-params for SiamFC-3S
% The ones for SiamFC (5 scales) are in params-5s.txt
p.numScale = 3;
p.scaleStep = 1.0375;
p.scalePenalty = 0.9745;
p.scaleLR = 0.59; % damping factor for scale update
p.responseUp = 16; % upsampling the small 17x17 response helps with the accuracy
p.windowing = 'cosine'; % to penalize large displacements
p.wInfluence = 0.176; % windowing influence (in convex sum)
p.net = '2016-08-17.net.mat';
%% execution, visualization, benchmark
p.video = 'vot15_bag';
p.visualization = false;
p.gpus = 1;
p.bbox_output = false;
p.fout = -1;
%% Params from the network architecture, have to be consistent with the training
p.exemplarSize = 127; % input z size
p.instanceSize = 255; % input x size (search region)
p.scoreSize = 17;
p.totalStride = 8;
p.contextAmount = 0.5; % context amount for the exemplar
p.subMean = false;
%% SiamFC prefix and ids
p.prefix_z = 'a_'; % used to identify the layers of the exemplar
p.prefix_x = 'b_'; % used to identify the layers of the instance
p.prefix_join = 'xcorr';
p.prefix_adj = 'adjust';
p.id_feat_z = 'a_feat';
p.id_score = 'score';
% Overwrite default parameters with varargin
p = vl_argparse(p, varargin);
% -------------------------------------------------------------------------------------------------
% Get environment-specific default paths.
p = env_paths_tracking(p);
1
% Load ImageNet Video statistics
if exist(p.stats_path,'file')
stats = load(p.stats_path);
else
warning('No stats found at %s', p.stats_path);
stats = [];
end
2
% Load two copies of the pre-trained network
net_z = load_pretrained([p.net_base_path p.net], p.gpus);
3
net_x = load_pretrained([p.net_base_path p.net], []);
4
p.seq_base_path
p.video
[imgFiles, targetPosition, targetSize] = load_video_info(p.seq_base_path, p.video);
nImgs = numel(imgFiles);
startFrame = 1;
5
% Divide the net in 2
% exemplar branch (used only once per video) computes features for the target
remove_layers_from_prefix(net_z, p.prefix_x);
remove_layers_from_prefix(net_z, p.prefix_join);
remove_layers_from_prefix(net_z, p.prefix_adj);
% instance branch computes features for search region x and cross-correlates with z features
6
remove_layers_from_prefix(net_x, p.prefix_z);
zFeatId = net_z.getVarIndex(p.id_feat_z);
scoreId = net_x.getVarIndex(p.id_score);
% get the first frame of the video
im = gpuArray(single(imgFiles{startFrame}));
% if grayscale repeat one channel to match filters size
if(size(im, 3)==1)
im = repmat(im, [1 1 3]);
end
% Init visualization
videoPlayer = [];
if p.visualization && isToolboxAvailable('Computer Vision System Toolbox')
videoPlayer = vision.VideoPlayer('Position', [100 100 [size(im,2), size(im,1)]+30]);
end
7
% get avg for padding
avgChans = gather([mean(mean(im(:,:,1))) mean(mean(im(:,:,2))) mean(mean(im(:,:,3)))]);
wc_z = targetSize(2) + p.contextAmount*sum(targetSize);
hc_z = targetSize(1) + p.contextAmount*sum(targetSize);
s_z = sqrt(wc_z*hc_z);
scale_z = p.exemplarSize / s_z;
% initialize the exemplar
[z_crop, ~] = get_subwindow_tracking(im, targetPosition, [p.exemplarSize p.exemplarSize], [round(s_z) round(s_z)], avgChans);
if p.subMean
z_crop = bsxfun(@minus, z_crop, reshape(stats.z.rgbMean, [1 1 3]));
end
d_search = (p.instanceSize - p.exemplarSize)/2;
pad = d_search/scale_z;
s_x = s_z + 2*pad;
% arbitrary scale saturation
min_s_x = 0.2*s_x;
max_s_x = 5*s_x;
switch p.windowing
case 'cosine'
window = single(hann(p.scoreSizep.responseUp) * hann(p.scoreSizep.responseUp)');
case 'uniform'
window = single(ones(p.scoreSizep.responseUp, p.scoreSizep.responseUp));
end
% make the window sum 1
window = window / sum(window(:));
scales = (p.scaleStep .^ ((ceil(p.numScale/2)-p.numScale) : floor(p.numScale/2)));
% evaluate the offline-trained network for exemplar z features
net_z.eval({'exemplar', z_crop});
z_features = net_z.vars(zFeatId).value;
z_features = repmat(z_features, [1 1 1 p.numScale]);
bboxes = zeros(nImgs, 4);
% start tracking
tic;
for i = startFrame:nImgs
i
if i>startFrame
% load new frame on GPU
im = gpuArray(single(imgFiles{i}));
% if grayscale repeat one channel to match filters size
if(size(im, 3)==1)
im = repmat(im, [1 1 3]);
end
scaledInstance = s_x .* scales;
scaledTarget = [targetSize(1) .* scales; targetSize(2) .* scales];
% extract scaled crops for search region x at previous target position
x_crops = make_scale_pyramid(im, targetPosition, scaledInstance, p.instanceSize, avgChans, stats, p);
% evaluate the offline-trained network for exemplar x features
[newTargetPosition, newScale] = tracker_eval(net_x, round(s_x), scoreId, z_features, x_crops, targetPosition, window, p);
targetPosition = gather(newTargetPosition);
% scale damping and saturation
s_x = max(min_s_x, min(max_s_x, (1-p.scaleLR)s_x + p.scaleLRscaledInstance(newScale)));
targetSize = (1-p.scaleLR)targetSize + p.scaleLR[scaledTarget(1,newScale) scaledTarget(2,newScale)];
else
% at the first frame output position and size passed as input (ground truth)
end
rectPosition = [targetPosition([2,1]) - targetSize([2,1])/2, targetSize([2,1])];
% output bbox in the original frame coordinates
oTargetPosition = targetPosition; % .* frameSize ./ newFrameSize;
oTargetSize = targetSize; % .* frameSize ./ newFrameSize;
bboxes(i, :) = [oTargetPosition([2,1]) - oTargetSize([2,1])/2, oTargetSize([2,1])];
% if p.visualization
if isempty(videoPlayer)
figure(1), imshow(im/255);
figure(1), rectangle('Position', rectPosition, 'LineWidth', 4, 'EdgeColor', 'y');
drawnow
fprintf('Frame %d\n', startFrame+i);
else
im = gather(im)/255;
im = insertShape(im, 'Rectangle', rectPosition, 'LineWidth', 4, 'Color', 'yellow');
% Display the annotated video frame using the video player object.
step(videoPlayer, im);
end
% end
if p.bbox_output
fprintf(p.fout,'%.2f,%.2f,%.2f,%.2f\n', bboxes(i, :));
end
end
bboxes = bboxes(startFrame : i, :);
end
A000`
