RFIW: Large-Scale Kinship Recognition Challenge
Over the past two years, I have worked as a research assistant on the Families in the Wild project, which consisted of building the largest kinship image dataset to date. We hope this will propel forward even more research in kinship related vision problems, since a large enough dataset that represents the complex hierarchical family relationships in the real world is necessary to make progress.
Recently, I assisted in organizing the first large-scale kinship recognition data challenge, Recognizing Families in the Wild in conjunction with ACM-MM 2017, which utilized our new dataset and provided an organized platform for participants to test their algorithms with respect to two tasks: a one-to-one Boolean verification problem (verify if two people are related), and a one-to-many classification problem (classify this person into a family). At the conference, we prepared a four hour workshop, where we presented the results of the challenge, along with talks from two keynote speakers, and two of the challenge participants. More information can be found here: https://web.northeastern.edu/smilelab/RFIW2017/ .
While helping organize a conference workshop and presenting is a valuable experience itself, it was only one day of the five day conference. This was my first time attending a research conference, and I believe it may be the catalyst to pursue a Ph.D. Between the keynotes and poster sessions, the developments I saw in deep-learning and multimedia are very exciting. It was also an interesting experience being surrounded by a community of people all working towards a similar goal in advancing computing in multimedia. I heard fascinating talks on topics such as reinforcement learning, person re-identification, and virtual reality in medicine.
When I started my undergraduate studies, I had little intention of even going to graduate school, but my experiences here have really sparked my interest for research in this field.
Conference Travel Fund Blog Post by Timothy Gillis
Candidate for Bachelor of Science in Computer Engineering and Computer Science ’20