Computer Vision Engineer Interview Questions
500
Computer Vision Engineer interview questions shared by candidates
Why does one use MSE as a measure of quality. What is the scientific/mathematical reason for the same?
3 Answers↳
Mean-Square error is an error metric for measuring image or video quality it is popular video and image quality metric because the analysis and mathematics is easier with this L2-Norm metric. Most video and image quality experts will agree that MSE is not a very good measure of perceptual video and image quality. Less
↳
The mathematical reasoning behind the MSE is as follows: For any real applications, noise in the readings or the labels is inevitable. We generally assume this noise follows Gaussian distribution and this holds perfectly well for most of the real applications. Considering 'e' follows gaussian distribution in y=f(x) + e and calculating the MLE, we get MSE which is also L2 distance. Note: Assuming some other noise distribution may lead to other MLE estimate which will not be MSE. Less
↳
MSE is used for understanding the weight of the errors in any model. This helps us understand model accuracy in a way that is helpful when choosing different types of models. Check out more answers on InterviewQuery.com Less
How would you code up a custom rectangle detector?
5 Answers↳
Hi... Could you plz explain what is a custom rectangle detector...I hav my interview with Amazon on monday ... plz help Less
↳
First of all, hough transformation can be used. just parametrize the representation for rectangle, but the parameter space is 4D. Second, line detection, followed by checking corner degree. In practice, I would use opencv's coutour fitting function to fit for quadrilateral, then check the angle. This works quite well. Less
↳
I would scan the image with a basic edge detection mask. Then I would scan to count the number of lines, keeping track of end points in pairs. Cases such as curves would also be found in this step and returned as not a rectangle. An analysis of basic trig with the points could then be performed in order to determine if the points form a rectangle. Once a rectangle is confirmed then you could use those points to display on the image the recognized rectangle. Or do whatever the reason for finding the rectangle was. Less
Given a 2D grid consisting of links and vacancies represented by a 2D array and assume you can move from either down (j,k)->(j+1,k) or to the right (j,k)->(j,k+1) unless there is a vacancy
4 Answers↳
Use recursion
↳
No well defined. Is the problem to find the shortest path to the right side or bottom or shortest path to the a particular location or to connect the most locations or what? Less
↳
given rectangle set (overlapped or not), calculate the overlapped areas. e.g., rect{ point upperleft; width, height; } Less
(1) Back-propagation; (2) overfit vs underfit; (3) implementation of one of the steps in Canny edge detection using graphs; (4) closed form formula to solution of linear regression; (5) write code for K nearest neighbor algorithm (6) vanishing gradient vs exploding gradients due to choice to relu() vs sigmoid()
3 Answers↳
Could answer most of it satisfactorily
↳
fliрkart.com
↳
flipkart
Question on finding all possible paths from start and end set
2 Answers↳
If you don’t mind sharing, when did you give the interview? Thanks a bunch!
↳
The key in these questions is to cover the fundamentals, and be ready for the back-and-forth with the interviewer. Might be worth doing a mock interview with one of the Magic Leap or ex-Magic Leap Computer Vision Engineer experts on Prepfully? They give real-world practice and guidance, which is pretty helpful. prepfully.com/practice-interviews Less
What is meant by virtual memory? Time complexity of insertion in Linked List vs. Array? Smoothing an image - what filter to use? Salt and pepper noise removal? Memory storage - heap vs stack?
2 Answers↳
Smoothing an image can be done by convolving with a kernel: 1- a matrix of ones- simple averaging. 2- convolving with a gaussian matrix which simply means the center pixel has more weight in averaging. Salt an pepper noise: with max - min or median filters. Less
↳
What is meant by virtual memory? virtual memory is a memory management technique that provides an "idealized abstraction of the storage resources that are actually available on a given machine" which "creates the illusion to users of a very large (main) memory". Time complexity of insertion in Linked List vs. Array? Linked list: O(1) Array: O(n) Memory storage - heap vs stack? Stack: linear data structure, high-speed access, no fragmentation, local variables only, variables can't be resized. Heap: hierarchical data structure, slow access time(compared to stack), memory can become fragmented, global variables, variables can be resized. Less
I had a take home assignment - which you code up and bring with you to the interview for a panel discussion and presentation. I had to implement a blob detection algorithm.
1 Answers↳
Can you use the OpenCV library?
Given a set of numbers, find the pair of numbers that has maximum difference and also the second number appears after the first one. Time complexity needs to be O(n).
1 Answers↳
#def max(a, b) (a>=b ? 1:2) #def min(a, b) (a= temp) { minimum = temp; imin = i+1; } } else { temp = A[i]; if (m >= temp) { minimum = temp; imin = i+1; } } } } Finally, the values in maximum and minimum will be the pair of numbers with the greatest difference between them, with their corresponding indices stored in imax and imin. The complexity of this implementation is O(n) because the loop goes through n-1 elements once, where n is the total number of elements. Strictly speaking, complexity is O(n-1). Less
A programming challenge was given, in Image processing applying filters. An assignment of Princeton University Archives. The solution was needed to be coded and submitted in 48 hours.
1 Answers↳
I implemented the solutions and submitted.
