Channel: Learn By Watch

Category: Education

Tags: cnnconvolutional neural network tutorialcnn machine learningconvolutional neural networkmachine learning coursecnn neural networkconvolution on rgb ledlenet-5 convolutional neural networkscnn machine learning tutorialconvolutional neural network pythonpooling convolutional neural networkscnn machineconvolutional neural network theorem problemsmaxpool in cnndifferent layers in cnncnn architecture

Description: A Convolutional Neural Network (ConvNet/CNN) is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various aspects/objects in the image and be able to differentiate one from the other. In this video you will learn about these topics: ● Recap of the convolution step ● Convolution on RGB - Until now we were looking at convolution on B/W images. But in the real world we have coloured images and here we discuss how we convolve on them. ● Convolution on RGB using multiple filters - While each filter can extract just one feature from the image, we can have multiple filters in each layer so that we can extract multiple features. ● Pooling layers - We looked at two types of pooling layers : ○ Max Pool - Most widely used. Takes the value of the maximum pixel ○ Average pool - Not used that much. Take the average of the pixels. ● Different layer in CNN - Convolution, pooling and fully connected ● LeNet-5 - Discussed the architecture of a standard CNN and understood how these standard models perform better than the models built from scratch. We can use these models in our own dataset by modifying them at the end.