Incoming EIC Editorial

AS an undergraduate engineering student at the University of Waterloo with a budding interest in computer vision, I discovered that by joining the IEEE as a student member, I could receive my own personal copy of TPAMI each month for barely the cost of a cup of coffee. TPAMI offered me an exciting window to the world of computer vision research, exposing me to its diversity of research topics, its leading researchers, and its international focal points of research activity. Thirty-four years later, during which one of my office walls has become largely yellow, TPAMI continues to be the computer vision community’s leading journal, led by a long line of remarkable EICs whose commitment to maintaining the excellence of the journal has been truly exceptional. It’s therefore a tremendous honor for me to be selected to serve as the next EIC of this illustrious journal. I would like to begin by expressing, on behalf of the entire TPAMI community, my sincerest thanks to David Forsyth for the outstanding leadership, vision, and commitment he has brought to TPAMI and to our community. During his watch as TPAMI EIC, the community has grown significantly in terms of the scope of our research, the number of active researchers in our community, the number of attendees at our conferences, and the engagement of industry. This growth posed many challenges to TPAMI, including how to handle the increase in both the volume and scope of submitted papers. Yet thanks to David’s expert guidance, the journal not only met these challenges, but grew stronger over his term, increasing its impact factor over his four-year term from 4.795 to 6.077. The journal now ranks 4th among all computer science journals according to Thomson Reuters Journal Citation Reports, up from 5th one year ago. The journal faces some important challenges ahead. Over David’s watch, our community has also become increasingly interdisciplinary, with machine learning emerging as our most prominent cognate discipline. In response to the journal’s growing scope and growing volume of submissions, David expanded the cohort of Associate Editors in Chief (AEICs) to its current size of six, including Amir Globerson, Dale Schuurmans, Kristin Grauman, Christoph Lampert, Kyoung Mu Lee, and Tinne Tuytelaars. They play a critically important role in the journal’s success, and I’m very much looking forward to working with each of them. This expanded pool of AEICs offers a number of important benefits, including allowing the workflow to be handled in shifts, providing a broad range of subdisciplinary expertise in paper assignment, and offering a valuable sounding board to the EIC for process changes and staffing decisions. As our field continues to expand into new directions and applications, it’s critical that we appoint AEICs (and AEs) that possess the expertise to handle papers in these emerging areas. Open Access (OA) continues to be an issue in our community. The Open Access Committee for CVPR/ICCV (of which I was a member) strongly endorsed the sponsorship/OA agreement reached between the IEEE Computer Society and the Computer Vision Foundation, addressing the need for open access while acknowledging the importance of our relationship with IEEE. The issue of OA with respect to journals, in general, and TPAMI, in particular, will continue to grow, and I view the 10-member TPAMI Advisory Board as playing a key role in determining how TPAMI responds to this critical issue. The Advisory Board has been a tremendously valuable resource to David and to the previous EICs, and its collective wisdom and experience will play a key role in guiding the journal through these and other issues that arise. Women continue to be underrepresented in the computer vision community. TPAMI can play an important role in addressing this gender gap by recruiting our field’s top women researchers into leadership positions, including Associate Editors in Chief (AEICs) and Associate Editors (AEs). David has worked very hard to recruit women into these roles, and this remains a priority for me as well. One of the greatest challenges facing TPAMI (and, in fact, all journals in CS) is the vision community’s adoption of our top conference (ICCV, CVPR, and ECCV) papers as not only archival but terminal (in the sense that many authors do not expand their conference papers into journal papers). While a conference paper represents an important milestone along the trajectory of a research project, it should not be the final milestone. I believe science is better served by researchers taking the feedback they receive at a conference and extending/improving their work in a more developed journal paper. A journal paper not only offers a more thorough review process, but its increased length provides additional space for the technical detail required for reproducibility. Of course, this is not a new challenge facing the journal, as past EICs have also faced this challenge. The mechanisms they’ve introduced, such as conference “best papers” fast-tracked as TPAMI journal papers, special issues on emerging topics, and reducing average-time-to-decision all represent effective incentives for TPAMI submission. To maximize the impact of our community’s research and to promote longer-term research, TPAMI must continue to explore creative ways to incentivize conference paper authors to take the time to further develop their