Computer Vision Software Engineer Interview Questions | Glassdoor

# Computer Vision Software Engineer Interview Questions

11

Computer vision software engineer interview questions shared by candidates

## Top Interview Questions

Sort: RelevancePopular Date

### Computer Vision/Machine Learning Software Engineer at Gauss Surgical was asked...

Jun 17, 2014
 Why does one use MSE as a measure of quality. What is the scientific/mathematical reason for the same?2 AnswersMean-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.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.

### Software Development Engineer (Computer Vision) at Amazon was asked...

Nov 22, 2012
 How would you code up a custom rectangle detector?4 AnswersI suggested something about matching intersecting hough lines.Hi... Could you plz explain what is a custom rectangle detector...I hav my interview with Amazon on monday ... plz helpBy a custom rectangle detector I mean how would you write your own function to detect rectangles. Your input would be the pixel data and your outputs would be something like the x,y locations of the rectangles.Show More ResponsesFirst 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.

### Software Engineer Computer Vision at Matterport was asked...

Mar 19, 2014
 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 vacancy4 AnswersUse recursionNo 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?given rectangle set (overlapped or not), calculate the overlapped areas. e.g., rect{ point upperleft; width, height; }Show More Responsesanother question: given rectangle set (overlapped or not), calculate the overlapped areas. e.g., rect{ point upperleft; width, height; }

### Computer Vision Software Engineer at TuSimple was asked...

Feb 22, 2017
 For the technical questions, he asked me about spanning tree and dynamic programming.Be the first to answer this question

### Computer Vision Software Engineer at Oculus VR was asked...

Sep 4, 2015
 I also had a phone screen where they asked me to write code on a shared screen. I think they asked me to remove duplicates from an array. The phone interview was quite straightforward and good.Be the first to answer this question

### 3D Computer Vision Software Engineer at VanGogh Imaging was asked...

May 22, 2014
 What is the difference between debug and release mode? Have you used revision/version control? How can you find 2nd highest number in an array in least time? How to eliminate corners in a bunch of 3D point cloud? How to find shape/object correspondences in 3D point cloud?Be the first to answer this question

### Computer Vision Software Engineer at Oculus VR was asked...

Sep 4, 2015
 How would you send an encoded message in an image. First encode it, then write code to decode it.2 AnswersMy initial idea was to use a digital watermark kind of technique using alteration of DCT coefficients and using that information to efficiently decode the message. However, there is a much better technique elaborated here - http://lucascozby.com/2015/09/15/encoding-messages-into-an-image/.I would always start with the simplest and fairly obvious answer; I would put the message to least significant bit layer of the image. The receiving end would only read the least significant bit layer of the image accordingly to decode the message. The least significant bit is invisible to the human eye. Test yourself by painting a rectangle with two reds next to each other of values (255,0,0) and (254,0,0). See if you can tell the difference.

### Computer Vision Software Engineer at Oculus VR was asked...

Sep 4, 2015
 Transpose a Matrix. First do it using extra memory, then do it in place.Be the first to answer this question

### Computer Vision Software Engineer at NVIDIA was asked...

Jun 18, 2014
 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?1 AnswerSmoothing 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.

### 3D Computer Vision Software Engineer at Ostendo Technologies was asked...

Jun 3, 2014
 Singleton pattern1 AnswerI always think that one is either good at Systems Engineering or at pure/generic Software. I was rejected at this interview despite the manager saying that most of the protoyping/research was performed on MATLAB. I thought the company didn't have a clear idea about what they wanted from the role/candidate. They were just matching skills that didn't matter too much to their product and the role.
110 of 11 Interview Questions

More