I work for a company that has several projects involving computer vision. The li
ID: 649380 • Letter: I
Question
I work for a company that has several projects involving computer vision. The literature rarely has papers on the exact problems we are solving. Furthermore, the state of the art in much of computer vision is often not reproducible or not ready for actual applications. Furthermore, the rate at which new papers come out in the areas we are interested in would make it very time consuming to digest most, let alone all of them.
On the other hand, we often get stuck and would like to get answers or at least inspiration from existing work.
Our goal in the end is to get to solutions as fast as possible. Does anyone have any guidance on how to balance time spent reading vs doing?
Explanation / Answer
Before trying to solve the problem yourself, it is always a good idea to search the literature, to see if somebody else has already solved it, or at least has worked on something similar. This is especially true in computer vision. A good strategy is to allocate some reasonable amount of time to read papers. If you find nothing useful in that time, then you give up and try to come up with your own solution.
Yes the number of computer vision papers published every year is staggering. So a good strategy is to try to keep up with a few top conferences, such as CVPR, ICCV, ECCV, and BMVC, and a few top journals, such as PAMI and IJCV.
Also, it is good to find researchers and research groups who are working in your particular area of computer vision, and follow their progress, and the papers which cite them.