The tutorial programme at SAMT 2006 is worth reading as a set rather than a menu. The first, The Reality of the Semantic Gap in Image and Video Analysis, takes the problem head-on and asks what is genuinely achievable today when you try to attach meaning to visual content. The second turns to Human Language Technology, where decades of work on extracting structure from text offer tools, and cautionary tales, for the media community. The third, Semantic Indexing and Retrieval of Video, follows the chain all the way to a working system that a user can query.

Put side by side, the three share a single conviction: progress comes from the seams. Image analysis gains from language technology; retrieval gains from both; and none of them gets far without a shared representation to hang the results on. That is the same message the keynotes deliver from the research frontier, restated here in a form you can take home and build with.

For newcomers the tutorials are also the gentlest way into the field. They assume curiosity rather than a decade of background, and they leave you able to read the main-track papers with a map in hand.