First part was talking about my experiences in terms of building backend pipelines, computer vision algorithms, 3D reconstruction concepts and general software engineering. The second part was more theoretical questions dealing with machine learning, mathematics, training models and linear algebra.
My flexibility with going for site deployments and testing.
Number of neurons in a neural network for a gray scale and RGB image?
Tensor Flow, About Dropouts, Model Training AWS use case -- with TWO faces and how to Detect the Face...AWS Beanstalk, AWS DynamoDB.... Computer Vision --- More about the works and Model Training, Metrics used, False Positive, False Negative....
Please descibe a large project you did and what kind of computer vision methods you used?
Implement a special type of matrix class using c++.
If I have scratches over a car and I need to use a camera to detect them, which AI model will you use to detect those scratches?
The technical questions in this first round were good, but easy. BUT They would have expected an unpaid programming assignment to be completed at a later stage (for which one week would have been available). I find this unacceptable.
What is my project experience?
7. Explain YOLO?
Viewing 561 - 570 interview questions