For pascal sentence dataset, simply :python multi_pascal.py

-h, --help: show this help message and exit

-m, --model: 1: ML-ELM, 2: HeMap (default: 1)

-j, --joint: joint method (cca, pcca) (default: cca)

--hidden: number of hidden units (default: 256)

--output: number of output units (default: 64)

-r, --reg: regularize parameter of feature extraction layers (default: 1)

-c, --cca: regularize parameter of cca (default: 1)

-l, --layers: the number of iteration of (ELM/HeMap + CCA) (default: 1)

-f, --file: filename (default: result.txt)

-q, --quiet: do not print to file

pascal dataset retrieval used the ML-ELM model in which the number of hidden units is 512 and the number of layers is 5 (2 iterations of ELM+CCA):

python multi_pascal.py --hidden 512 -l 2