Computer vision is a challenging field with tremendous promise. With greatly improved processor speeds and memory capacity, the promise of computer vision attracts attention, yet its challenges remain. A regular reading group fulfills several important purposes:

The overall goal is to understand and overcome the challenges of computer vision through regular discussion.


While the reading group is organized by and hosted at HP Labs, anyone interested in computer vision may attend.

The discussions are often focused around a single publicly-available document, such as a conference paper or book chapter. No HP intellectual property is discussed at our meetings.


In general, we choose topics based on the interests of the participants. Because we meet once per week, we are able to cover many subjects in a short amount of time. To keep up this pace, we limit ourselves to one paper per session. Areas of interest expressed thus far (in no particular order, and in rough grouping):


We have very flexible guidelines for each discussion. The format can range from a whiteboard session to a prepared presentation, and the content source can range from a book chapter to your working research ideas that you would like to vet in front of an audience. Most importantly, it is not even necessary for the leader to fully understand the topic at hand; instead, the discussion leader briefly summarizes the topic and directs the discussion.

Discussion questions to keep in mind:

Journals and Conferences

CVPR - Computer Vision and Pattern Recognition
ECCV - European Conference on Computer Vision
ICCV - International Conference on Computer Vision
IJCV - International Journal of Computer Vision
PAMI - IEEE Transactions on Pattern Analysis and Machine Intelligence
VG - International Workshop on Volume Graphics