Google doesn`t consider CSS ( web layout ) elements as duplicate content - they are however extremely sensitive to text and links. Embedded media is quite often laid on a Jquery or Javascript code, so it cant read that..hyperlinked images however may look suspicious because of the text in the images especially if they have same alt attributes and links.. I guess as long as you do not have similar patterns of text and links, it wouldnt be comsidered a duplicate. .. One way to compare them is by trying to read the text-only version of your webpages as this is what the crawler reads and indexes.

There's hardly any text on the Contact Us page. All it has is the form for people to contact.

Google Panda filter out only duplicate and syndicated contents.

Google can read only text. Carolers can not read images, videos.

They can't really track this kind of media, so it is mostly texts and links. When duplicate content is found, Google usually punishes the last website that posted it, rather than punishing both people.