SIGGRAPH

I have often been asked for suggestions on how to manage one's time at SIGGRAPH, given that there are always so many things going on at once. My answer is always the same. I prefer to immerse myself in as many technical papers sessions as possible so that I can hear about all of the wonderful new research directly from the authors. Of course you can read papers before or after the conference, but I almost always find that the best introduction to the ideas is the author's own words and images from their talks. Part of what makes the SIGGRAPH conference so special is the care that the authors put into their presentations, making the talks a joy to attend. This year the papers program is filled with exciting papers that span the entire range of topics in graphics. I hope that you will take the opportunity to attend many of the talks so that you can enjoy these new research contributions more fully.

There were 518 papers submitted to the ACM SIGGRAPH 2008 Technical Papers Program, and 90 of these papers were accepted to the conference. These paper counts have more meaning when compared to previous conferences, and I refer you to the table and graph at the end of this preface to see the submission and acceptance rates for earlier SIGGRAPH conferences. What you will find is that the number of submissions this year took a big jump from previous years, but that the acceptance rate is in line with previous years.

The biggest change to the technical papers program this year is the introduction of TOG presentations to SIGGRAPH. Every author of a paper appearing this year in the ACM Transactions on Graphics (TOG) has the opportunity to give a full presentation of their work in a SIGGRAPH technical papers session. As I write this, the October 2008 issue of TOG has yet to be compiled, so we do not know the exact number of TOG presentations that will be at the conference. Our best guess is that roughly 30 such TOG presentations will be given, spread over seven or eight different papers sessions. This important change to the conference was the result of long deliberations by a working group that was led by Rob Cook, and that also included Adam Finkelstein, John Hart, Jessica Hodgins, Holly Rushmeier and myself.

In addition to TOG presentations at SIGGRAPH, there is another way that TOG and SIGGRAPH are tied together. For several years now, in addition to the choices of "accept" or "reject", there has been a third option in the review process for papers submitted to SIGGRAPH. If the reviewers think that the content of a paper is SIGGRAPH quality, but that the paper needs more revisions than time permits in the conference review cycle, such a paper can be designated as "reject and refer to TOG." Then, if the authors wish, their paper can be brought into the TOG review process, carrying with it the SIGGRAPH reviews and the ongoing input from the SIGGRAPH papers committee members that made the referral. This year, 24 papers were referred to TOG in this manner.

One of the most important responsibilities of the SIGGRAPH papers chair is to seek out ways in which the papers review process can be improved. The SIGGRAPH review process is considered by many to be a model of fairness and care. I have talked to papers chairs from other computer science conferences that have deliberately borrowed ideas from the SIGGRAPH review process to improve their own conference. One of the most frequent comments that I hear from first-time papers committee members is how much better they feel about the review process once they have seen what the face-to-face papers committee meeting is like. As good as the review process is, however, there is always room for improvements. A concern that is sometimes voiced is that papers committee members have too much control over the decision-making process for papers, in contrast to the tertiary reviewers who do not attend the PC meeting. To help address this issue, all tertiary reviewers this year were able to post to the public bulletin-board during the rebuttal phase of reviews. In addition, I asked all of the reviewers (including tertiaries) to try to reach a reviewer consensus for each paper prior to the papers committee meeting, via discussions on the private bulletin-boards. I view this as the logical progression that has been taking place over the last several years of including tertiary reviewers more in the decision process.