I did the program in Python quickly using 2D array search and if-else conditioning. The interviewer agreed with my solution. However, he did mention that if I can do it in C++, it will be even better. Menos
I described my understanding of a feed-forward neural network and later I enhance the flexibility and robustness of the framework, and apply it to some classification problem that in my daily work. Menos
Asked about experience in linguistics and computer science, specifically with pattern recognition tools and models. Asked for details about relevant experience (in previous work, in school, and in outside projects) and for specifics about what I worked on and what tools I used.
1. Describe the research activities and projects that I have worked so far in speech.
2. What is my experience in speaker diarisation?
3. How do I address the problem of speech overlap in diarisation?
4. How familiar I am with speech tools
5. Describe the working of E2E and Chainmodels in ASR
6. Explain bias-variance trade-off