|
JournalFire makes it easy to share and discuss journal articles. Learn more...
|
We are preparing a journal version of this paper. Please, let us have your comments!
We are preparing a journal version of this paper. Please, let us have your comments!
How does this article relate to Ramanans?
This would be a good paper to discuss jointly. The main advance w.r. to the `classic' constellation model here seems to be the idea of training parts differentially (category vs. background). This is a clever idea. I am suspicious of training differentially category vs. category because this makes incremental training much more difficult. Pietro
I added this article to test the system.
John:
Gilles is telling me that you are planning to present journalfire to the Caltech community sometime in September. This is a great idea. Please do advertise on cns-interest.
My impression of journalfire is that it is trying to carry out a similar mission as the movie comments you find in imdb.com
It would be important to sort out all the ways in which it could be used:
- journal club within a research group or a local community
- journal club in a distributed community
- helping me find the best paper on a certain topic
- helping authors respond to common misconceptions about their papers
- helping scientific communities to quickly pinpoint problems with papers
- helping sync ... read more
