2016-11-02 03:16:56 -0600 | answered a question | Canny Edge Detection- Non Maximum Surpression Implementation Maybe the issue is with how you're calculating the Sobel gradients. Try this . where ksize is the size of the Sobel kernel. (in your case, ksize = 3) |
2016-10-27 00:54:49 -0600 | commented answer | How to classify an object that does not belong to the trained classess using MLP I wouldn't classify it as an overfitting, but yes, the consequences are similar. The only thing I can think of right now is apply your feature extraction algorithm and then instead of passing it through an ANN right away, pass it through a thresholding algorithm. Use the existing classes to form sort of a D-dimensional sphere around a center point (which can be the feature average of the known classes). If the data point falls outside this circle then store it as 'unknown'. Once you have enough in the unknown pool, you can retrain the NN with x+1 classes and update the thresholding algorithm with the new data. |
2016-10-26 01:28:51 -0600 | answered a question | How to classify an object that does not belong to the trained classess using MLP ANNs don't work like that. If a certain item is not in your training set, and you pass that item as an input to your network then it will classify that item to the closest possible set (ideally). One possible solution is using a regression model, that way you can possibly use the classification error to penalise this misclassification. But this method is pretty dangerous, and not recommended. |
2016-10-26 01:11:52 -0600 | asked a question | Python and C++ are giving different results for Hough Circle detection Hi all, So I have implemented the same code in both C++ and Python. The code does nothing much, just reads a grayscale image, implements the hough circle detection algorithm and gives the output. Code in C++: And for Python: These codes are quite similar, but the results are different for C++ and Python - Output from C++: (more) |