# Tag Info

0

Traditional CNNs used for image classification (and related tasks) are composed of 1 or more fully connected layers (FCs), after the convolutional and pooling layers, which take as input the features extracted from the convolutional and pooling layers, in order to perform classification or regression. One problem with FCs in CNNs is that the number of ...

2

It takes a little bit of time to fully understand the 2D convolution/cross-correlation and to relate it to the usual diagrams of the convolution operation, so, before addressing your questions, let me first try to break the definition of the 2D cross-correlation down, from the left to right. $$S(i,j) =(K*I)(i,j) = \sum_m \sum_n I(i+m, j+n)K(m,n) \label{1}\... 1 but I have been told that neural networks aren't made to predict values in that way, they really are best suited for classification into discrete classes I don't agree with this statement. I already trained many CNNs for regressions tasks where a continous output is trained and they generally perform very well. I think the general "advantage" for ... 0 I think U are looking for PLSA for PLSA either U find out those topics(catogeries) with EM or NNMF Personally I recommend NNMF or u can use LDA which is Bayesian version of PLSA here is code for PLSA: https://github.com/Man-ash/Probabilistic-Latent-semantic-analysis which use NNMF for EM method I code it by myself but i am not sure if it is right https://... 0 The other answer gives a good overview of the differences between MLPs and CNNs, and it includes 2 diagrams that attempt to illustrate the main differences between MLPs and CNNs, i.e. sparse connectivity and weight sharing. However, these diagrams do not clarify what a neuron in a CNN could be. A better diagram, which illustrates what a neuron is in a CNN, ... 0 There are many resources that answer your question, but, given that you're apparently new to machine learning (ML), deep learning (DL), and neural networks (NN), let me provide a simple answer that should clarify your doubts. The term weight in the context of ML, DL, and NN is a synonym for parameter (sometimes, in some contexts, such as linear regression, ... 1 If you do not specify an activation for a layer you are effectively creating a linear transformation through that layer. From the documentation: activation: Activation function to use. If you don't specify anything, no activation is applied (see keras.activations). 1 Create two different optimizers and split the subnets' parameters into either with different lrs. You will have to call optimizer1.step(), optimizer2.step() with a single backward() call 2 It sounds like what you're suggesting is similar to what is done in methods that use a planner. These methods looks to learn the dynamics of the MDP to use to plan during training; that is they want to be able to learn the transition probabilities p(s'| s, a). In this paper that I read recently they note that learning to predict environment dynamics when ... 1 The application of 1 kernel (aka filter) to an input (with a 2d convolution) is a matrix (a 2d array), which is often known as a feature map (aka activation map). The application of k kernels to the same input is a 3d array (sometimes called tensor, though this may not be exactly correct, or 3d volume) with depth k, i.e. you have k concatenated feature ... 0 Similar to other answers, I don't know Matlab that well but you could try the following steps to debug your problem. Make sure you can overfit to a single instance from your dataset, pull out a single image with a good amount of true positives in it. Duplicate that images B times (where B = Batch Size) and then try to train your network with only that small ... 0 I have never used MATLAB for ml before, so it is difficult for me to understand all your code. My first association to your problem is class imabalance. Since you seem to have got a handle on that, the problem could be dying ReLU or bloated activations. To check if the ReLU is dying, you could look at the activations of the early layers of your network. If ... 3 The point is that in the expansive path you have two forms of information: the information from the contracting path, which includes all high-level features extracted from the original image. the information from the skip-connections, which copy a cropped version of the feature maps in the contracting path. Because, as we go forward through the expansive ... 0 Noise vector (batch_size, 100, 1, 1) is deconvoloved with filter_1 (100, 4, 4). Result is feature_map_1 (1, 4, 4). And since there are 512 filters, so there will be 512 feature maps. Output shape will be (batch_size, 512, 4, 4). I think you need better undersanting for convolutional calculations in general. In this stack it was explained very well. 0 Here I am just trying to simplify what @user3667125 already said uses math arguments Say we have a cost function J(x, y; F(\cdot; \Theta)) which regards training a NN F(\cdot, \Theta) with x input and y expected output Gradient descent tells us how to upgrade each \theta_{i} \in \Theta and it is$$ \theta_{i}(t+1) = \theta_{i}(t) - \alpha \frac{\...

2

You can think about the problem in the following way (without padding, as the padding case is a simple extension of base case with $\tilde{W}:=W + 2P$). You want to know how many windows are necessary to cover an image of size $W$, given a window of size $K$ and stride $S$. So your image is a vector with indices $1, 2\dots, W$; as you put the first window on ...

1

Yes, your thought experiment is correct, and the concept is known as breaking the symmetry. This is why biases can be initialized to $0$ (bias initialization doesn't matter), but weights should be randomly initialized to different numbers -- to break the symmetry. Otherwise, if not, the network will function as if it has $n-1$ filters (or however many ...

0

Spectral Graph Convolution We use the Convolution Theorem to define convolution for graphs. The Convolution Theorem states that the Fourier transform of the convolution of two functions is the pointwise product of their Fourier transforms: $$\mathcal{F}(w*h) = \mathcal{F}(w) \odot \mathcal{F}(h) \tag{1}\label{1}$$  w * h = \mathcal{F}^{-1}(\mathcal{F}(w)\...

1

Look at spatio-temporal CNNs which extend the image-based CNN in 2D to 3D to handle time. These are commonly used to detect or classify action in a video. People have used them to identify specific actions in various sports such as kicking a soccer ball, throwing a baseball or dribbling a basketball. They have been used to identify fire, smoke, deep fakes,...

Top 50 recent answers are included