Unable to load video dataset

Hello guys. Does anyone familiar in using DTF file extension, which is used for action recognition coding. But in my situation, i do have the AVI file for datasets. Do i need to convert this AVI file to the DTF file extension ?

unzip_cmd=sprintf('gunzip %s',dtf_file); % Suppose dtf files have suffix .gz
system(unzip_cmd);
unzip_dtf_file=regexprep(dtf_file,'\.gz$',''); % remove suffix .gz
x=load(unzip_dtf_file);
zip_cmd=sprintf('gzip -f %s',unzip_dtf_file); % Suppose dtf files have suffix .gz
system(zip_cmd);

hog_range=params.feat_start:params.feat_start+params.feat_len_map('HOG')-1;
hof_range=hog_range(end)+1:hog_range(end)+params.feat_len_map('HOF');
mbhx_range=hof_range(end)+1:hof_range(end)+params.feat_len_map('MBHx');
mbhy_range=mbhx_range(end)+1:mbhx_range(end)+params.feat_len_map('MBHy');

if num_feats<0 % To use all the DTF fatures, set num_feats to a negative number
	num_feats=size(x,1);
end

if size(x,1)<=num_feats
	idx=1:size(x,1); % randomly subsampling
else
	idx=randperm(size(x,1),num_feats); % randomly subsampling
	%idx=floor(linspace(1,size(x,1),num_feats)); % linearly subsampling
end
HOG=x(idx,hog_range)';
HOF=x(idx,hof_range)';
MBHx=x(idx,mbhx_range)';
MBHy=x(idx,mbhy_range)';

end

In this action recognition coding which is written by Bo Yang, he used a compressed file of UCF101 (in a .gz file extension, compatible for linux). So after he uncompressed the file, he will obtained the DTF file then load it (x = load(unzip_dtf_file)). From my studies, DTF file is a file extension for dataset.

But in my situation, I'm using Windows. So that I have the uncompressed action video datasets (UCF101) in AVI file. But then I don't have any idea what to do on the 'load' part. Do i need to convert my AVI file to DTF file ? or I can extract directly the dataset properties from my AVI file?

Can anyone help me? I really hope that anyone can help me to solve my confusion.

[Mohd Shah Rizal SAMSUDIN]

Hello Muhammad, the dtf_file above was written in binary file. You need to go through the process of low-level feature extraction that will return into binary file. Unfortunately, you cannot use matlab for that early process. However I already shared all the process from scratch how to run IDT by using Virtual Box+Ubuntu. The link is available here (https://baitulaadiyat.blogspot.com/2020/07/dense-trajectory-and-improve-dense.html).

After you successfully, get the binary files, then you can proceed on matlab process. Many Thanks.