¡Hola, Mundo!

R-CNN(Regions with Convolutional Neural Networks features) The first model to apply deep learning of object detection. Region Proposal : With Selective search to identify a number of bounding-box object region candidates (“regions of interest”) CNN : Extracts feature vectors via CNN SVM : Classify the feature vectors with SVM Bounding Box Regression : Apply the Bounding Box Regression to predict..

Non-maximum Suppression (NMS) A technique that’s used after the region proposal step to eliminate duplicate bounding boxes and select the most relevant ones. The idea behind NMS is straightforward. It works by comparing the confidence scores of the proposed bounding boxes and eliminating the ones that overlap significantly with a higher-scoring bounding box. https://www.jeremyjordan.me/object-de..

Bounding Box Regression It is a metric for measuring how well predicted bounding boxes captures objects. This is done by calculating the mean-squared error (MSE) between the coordinates of predicted bounding boxes and the ground truth bounding boxes. Calculation The MSE between two bounding boxes, P and Q is given by: Where 4 comes from the fact that we want the mean of 4 coordinates (x1, y1, x2..

RoI(Region of Interest) A proposed region from the original image. In other words, finding interesting areas. We need to find interesting parts of the image, where the object is. Feature extraction Our model takes an image input of size 512x512x3 (width x height x RGB) and VGG16 is extracting it into a 16x16x512 feature map. If you look at the output matrix you should notice that it’s width and ..

Exhaustive Search It aims at finding all the possible locations by systematically enumerating all the possible candidates. Segmetation It uses the image structure to cluster pixels into different regions. Selective Search A method of improving Exhaustive Search by combining Segmetation. One of the method of Region proposal which has Sliding window and Selective search. https://medium.com/@rishu_..