What is the difference between a convolutional neural network and a . . . A CNN, in specific, has one or more layers of convolution units A convolution unit receives its input from multiple units from the previous layer which together create a proximity Therefore, the input units (that form a small neighborhood) share their weights The convolution units (as well as pooling units) are especially beneficial as:
What are the features get from a feature extraction using a CNN? By accessing these high-level features, you essentially have a more compact and meaningful representation of what the image represents (based always on the classes that the CNN has been trained on) By visualizing the activations of these layers we can take a look on what these high-level features look like
machine learning - What is a fully convolution network? - Artificial . . . A fully convolutional network is achieved by replacing the parameter-rich fully connected layers in standard CNN architectures by convolutional layers with $1 \times 1$ kernels I have two questions What is meant by parameter-rich? Is it called parameter rich because the fully connected layers pass on parameters without any kind of "spatial
In a CNN, does each new filter have different weights for each input . . . Typically for a CNN architecture, in a single filter as described by your number_of_filters parameter, there is one 2D kernel per input channel There are input_channels * number_of_filters sets of weights, each of which describe a convolution kernel So the diagrams showing one set of weights per input channel for each filter are correct
What is the fundamental difference between CNN and RNN? A CNN will learn to recognize patterns across space while RNN is useful for solving temporal data problems CNNs have become the go-to method for solving any image data challenge while RNN is used for ideal for text and speech analysis
Extract features with CNN and pass as sequence to RNN $\begingroup$ But if you have separate CNN to extract features, you can extract features for last 5 frames and then pass these features to RNN And then you do CNN part for 6th frame and you pass the features from 2,3,4,5,6 frames to RNN which is better The task I want to do is autonomous driving using sequences of images
Reduce receptive field size of CNN while keeping its capacity? One way to keep the capacity while reducing the receptive field size is to add 1x1 conv layers instead of 3x3 (I did so within the DenseBlocks, there the first layer is a 3x3 conv and now followed by 4 times a 1x1 conv layer instead of the original 3x3 convs (which increase the receptive field))
When training a CNN, what are the hyperparameters to tune first? Firstly when you say an object detection CNN, there are a huge number of model architectures available Considering that you have narrowed down on your model architecture a CNN will have a few common layers like the ones below with hyperparameters you can tweak: Convolution Layer:- number of kernels, kernel size, stride length, padding