Two more MM Grand challenges have been included in ACM MM conference. Bothe these are related to new things happening — one is related to social media and the other is related to defining events and tracking them in unstructured video. Both are challenging and interesting topics.
On one hand, I am very happy that academic community is identifying such challenges and trying to rally researchers around those. On the other hand, I am wondering about the strategy to select Grand Challenges and encourage community to work on them followed this year. By selecting so many Grand Challenges for such a small community — no challenge really remains a Grand Challenge. These Grand Challenges have mostly become research projects sponsored by special groups, particularly industry, who could get some ideas and research done without spending there resources. In many cases, these challenges are not even articulated with well defined goals.
In many cases one has good intentions to define and develop a focused approach, but pressures from different groups end up pulling and pushing it so much that in place of a focused approach one ends up adopting a shot-gun approach. It appears that that is what happened to MM Grand Challenge 2009 — in place of a Grand Challenge, I see total 10 challenges listed. And I know that researchers in my own group are struggling to decide which one they want to participate in. And I am afraid that in place of working on a grand challenge, they are wasting their time analyzing and reanlyzing these challenges to decide which one they should work.
I hope that MM Grand Challenge for ACM MM 2010 will have just one GRAND CHALLENGE.
Ramesh, this is a really good point but I have to ask:
One? There are like 4 tracks at ACM MM: Content, Systems, Applications, and Interactive Arts. How could one challenge give everyone something to chew on? Unless that one challenge was faceted well enough to suit. And if that is the case, are we all really working on the same problem?
Interestingly, David, none of the grand challenges seems to be directly tied to a track — they do have strong flavors. Most of them are biased towards content.
Even if we want to have grand challenges related to track — then lets call them that and have one for each track.