The job was about developing computer vision algorithms code for some specific hardware. I've been given a notebook to write some code in Python solving simple computer vision basic problems, The difficulty was to use Numpy features to reach maximum code efficiency and avoid using regular cycles etc,
theory of deep learning activation functions
1. Previous task 2. Python problem/code to be explained in front of the panel. Mainly based on generator, inheritance, polymorphism.
Fai una descrizione generale del tuo curriculum
The position was of a senior technical consultant that aided the team in delivering a sequence of prototypes based on new technologies. Although this full-time position lasted over 2.5 years it was from the very beginning a good understanding about a tentative nature of this arrangement. Most of the questions were very technology specific. One must also understand that IP was being filed so no direct questions were asked due to the sensitivity of the legal process in progress. So typical questions in this category are: 1. Given the physical principle used, can (my) device be expandable to larger format? 2. How this can be compared to the work done by so and so? 3. Were I involved in standard committees? (yes) 4. Possible applications for the physical effect xyz? 5. Issues in software implementation of the application function that is based the principle under discussion?
Depth reconstruction task from series of stereo images using 2 proposed algorithms. Compare to GT and estimate which algorithm performed better
Do you take ZZZ into account when you evaluate the performance?
Why you choose XXX method in your master thesis but not YYY?
Exibindo 1 a 10 perguntas de entrevista