A bit of background: I am working in a R&D department of quite a big name in the IT industry, and we are trying, clashing and combining different ideas and cutting-edge inventions to tackle our engineering problem.
Now the problem itself: There is an excellent paper from CVPR 2017 that I want to try for my project. However, the authors did not publish nor their code, nor their trained FCNN (Fully-Convolutional Neural Network). Now, the project that I am working on is only on the infancy stage, meaning that we need to make some Proof-of-Concept product as soon as possible. Obviously, for this stage, training a deep CNN is a relatively very time-consuming and expensive process, so ideally I would want to ask the authors for the trained model of their FCNN. Of course, if the concept works, we will re-train and fine-tune the whole thing for our specific task, so the author's CNN will not end up in the final product.
Question: How to correctly ask the authors for it? Or is it correct to ask researchers for their results for a project at all?