Content Based Video Retrieval - Event based Semantic Detection

Just a little bit more until the end of my 500 word PHD draft mini proposal

Darn... I can't think anymore. And to make things worse, the UPM connection won't let me go into Google....

Anyway, I'm trying to understand this Video Retrieval thinggy.
From what I read about Content Based Video Retreival, it seems to saying that... in the case of video, you can't use the old Content Based Image Retrieval technique such as low-level feature (such as colors, edges, textures and shape) extraction....
But instead model some sort of thing... that will enable the low-level fuatures to be able to map to the semantic concepts of that particular video!

Huh?

What did I just say?

I think I got all my sentences mixed up. It's 1:30 am and I'm getting pissed.

Anywayz, hope I can complete everything by tomorrow.
Plus I got my Digital Publishing paper to invigilate in another 6 hours plus! LOL!

Just to post!!!

Thank you oh patient readers...

PS:- Is this considered a rant?