This kind of site tries to make it difficult not to use a graphical web browser, because if you use wget you'll be missing all these ads that pay for the bandwidth.
Some sites don't make advanced checks and can be tricked easily: tell wget to pretend that it's really Mozilla and that it's coming from the download site.
wget --user-agent='Mozilla/5.0 (Windows NT 6.0) Gecko/20100101 Firefox/14.0.1' \
Most sites that check let you get away with
--referer set to the URL of the file you're downloading.
With some sites, you might need to export the web browser cookies and pass
--load-cookies to wget; at this point using wget starts to be more work than a manual download. The ultimate method, which can work all the time but requires some coding for each site, is to use a browser automation framework such as Selenium or Watir.
See also Does anybody here have experience in automating some tasks in web applications using curl?