Non-maximum suppression is used in CV in object detection where multiple probabilities that are above the given threshold are detected(classification result). Thus non-maximum suppression can help to suppress all the gradient values (by setting them to 0) except the local maxima classification result. If the strength of the current pixel/elemntContinue Reading
Pooling is a form of non-linear down-sampling(image reduction) with an aggregation function. Max pooling is the most common – It partitions the input image into a set of non-overlapping rectangles(N x N) and, for each such sub-region, outputs the maximum, thus, practically extracting the most significant features. The non-overlapping poolingContinue Reading
Artificial neural network is a structure composed of artificial neurons or nodes which are connected one to another. An artificial neuron receives one or more inputs that are separately weighted, and sums them all to produce an output. The output is passed through a non-linear function, known as an activationContinue Reading
Loss functions in ML(Machine Learning) are computational functions. These functions evaluate the price paid for inaccuracy of predictions in classification problems. Therefore, thier output should be minimized. The Loss Function used in Logistic Regression is cross entropy function: L(yˆ,y) = -(y*log(yˆ) + (1-y)*log(1-yˆ)) Where y is the ground truth outcomeContinue Reading
Logistic Regression can help predict specific values or categories with probability scores underlying the model’s classifications.Continue Reading
Similarity Measure The Jaccard index/coefficient, aka Intersection over Union is used in computer vision as one of the similarity measures for object detection on images. Formula The Jaccard similarity coefficient,aka Intersection over Union (IOU) score: Binary classification Formula The same score above can be described as binary classification as well to get anContinue Reading
A classifier is a method for determining the most likely class of an unknown object or event based on a number of instances of each of the classes (AKA the training set). There are two main ways to determine this, either by prior knowledge of what the output valuesContinue Reading
Depth of Field – Moving a little bit out of focus.. so we can see only the picture without the background –> depends on the size of the aperture – small aperture = the spread is less more focused background and vice versa. Field of View – increasing theContinue Reading
Yair Shinar for Clarity and Solutions