since no loops and no bin2dec, this work could be done using the definition of binary number and vecterization operation. Take binary number 101 for example:
a = 101;
b = str2num(num2str(a)');
c = 2.^(2:-1:0)*b;%get the decimal number here
it is easy to handle the binary number with 4 or more digits and even numbers with points.
It seems that this is a question in MATLAB Cody.....