Questions tagged [convolutional-neural-networks]

For questions about convolutional neural networks, also known as CNN or ConvNet.

Filter by
Sorted by
Tagged with
90
votes
7answers
16k views

Do scientists know what is happening inside artificial neural networks?

Do scientists or research experts know from the kitchen what is happening inside complex "deep" neural network with at least millions of connections firing at an instant? Do they understand ...
82
votes
9answers
6k views

How is it possible that deep neural networks are so easily fooled?

The following page/study demonstrates that the deep neural networks are easily fooled by giving high confidence predictions for unrecognisable images, e.g. How this is possible? Can you please ...
77
votes
3answers
55k views

How can neural networks deal with varying input sizes?

As far as I can tell, neural networks have a fixed number of neurons in the input layer. If neural networks are used in a context like NLP, sentences or blocks of text of varying sizes are fed to a ...
57
votes
10answers
44k views

In a CNN, does each new filter have different weights for each input channel, or are the same weights of each filter used across input channels?

My understanding is that the convolutional layer of a convolutional neural network has four dimensions: ...
36
votes
5answers
67k views

What is the difference between a convolutional neural network and a regular neural network?

I've seen these terms thrown around this site a lot, specifically in the tags convolutional-neural-networks and neural-networks. I know that a neural network is a system based loosely on the human ...
35
votes
6answers
10k views

Why do CNN's sometimes make highly confident mistakes, and how can one combat this problem?

I trained a simple CNN on the MNIST database of handwritten digits to 99% accuracy. I'm feeding in a bunch of handwritten digits, and non-digits from a document. I want the CNN to report errors, so I ...
26
votes
5answers
31k views

How can I deal with images of variable dimensions when doing image segmentation?

I'm facing the problem of having images of different dimensions as inputs in a segmentation task. Note that the images do not even have the same aspect ratio. One common approach that I found in ...
25
votes
4answers
415 views

Is the pattern recognition capability of CNNs limited to image processing?

Can a Convolutional Neural Network be used for pattern recognition in problem domains without image data? For example, by representing abstract data in an image-like format with spatial relations? ...
23
votes
3answers
39k views

How do I handle large images when training a CNN?

Suppose that I have 10K images of sizes $2400 \times 2400$ to train a CNN. How do I handle such large image sizes without downsampling? Here are a few more specific questions. Are there any ...
13
votes
2answers
6k views

Which layer in a CNN consumes more training time: convolution layers or fully connected layers?

In a convolutional neural network, which layer consumes more training time: convolution layers or fully connected layers? We can take AlexNet architecture to understand this. I want to see the time ...
12
votes
2answers
13k views

What are bottleneck features?

In the blog post Building powerful image classification models using very little data, bottleneck features are mentioned. What are the bottleneck features? Do they change with the architecture that is ...
12
votes
1answer
4k views

What is the difference between a receptive field and a feature map?

In a CNN, the receptive field is the portion of the image used to compute the filter's output. But one filter's output (which is also called a "feature map") is the next filter's input. What's the ...
11
votes
5answers
8k views

What is the fundamental difference between CNN and RNN?

What is the fundamental difference between convolutional neural networks and recurrent neural networks? Where are they applied?
11
votes
4answers
1k views

What are the models that have the potential to replace neural networks in the near future?

Are there possible models that have the potential to replace neural networks in the near future? And do we even need that? What is the worst thing about using neural networks in terms of efficiency?
11
votes
3answers
10k views

Is it difficult to learn the rotated bounding box for a (rotated) object?

I have checked out many methods and papers, like YOLO, SSD, etc., with good results in detecting a rectangular box around an object, However, I could not find any paper that shows a method that learns ...
10
votes
1answer
402 views

Can I start with perfect discriminator in GAN?

In many implementations/tutorials of GANs that I've seen so far (e.g. this), the generator and discriminator start with no prior knowledge. They continuously improve their performance with training. ...
10
votes
2answers
1k views

Do deep learning algorithms represent ensemble-based methods?

According to the Wikipedia article on deep learning: Deep learning is a branch of machine learning based on a set of algorithms that attempt to model high-level abstractions in data by using a deep ...
10
votes
2answers
1k views

What benefits can be got by applying Graph Convolutional Neural Network instead of ordinary CNN?

What benefits can we got by applying Graph Convolutional Neural Network instead of ordinary CNN? I mean if we can solve a problem by CNN, what is the reason should we convert to Graph Convolutional ...
9
votes
2answers
4k views

How can I use neural networks for detecting TV channel logos in video frames?

I am trying to detect a TV channel logo inside a video file. So, simply, given an input .mp4 video, detect if it has that logo present in a specific frame, say the ...
9
votes
1answer
361 views

How much of a problem is white noise for the real-world usage of a DNN?

I read that deep neural networks can be relatively easily fooled (link) to give high confidence in recognition of synthetic/artificial images that are completely (or at least mostly) out of the ...
9
votes
2answers
10k views

What is the concept of channels in CNNs?

I am trying to understand what channels mean in convolutional neural networks. When working with grayscale and colored images, I understand that the number of channels is set to 1 and 3 (in the first ...
9
votes
1answer
3k views

In YOLO, what exactly do the values associated with each anchor box represent?

I'm going through Andrew NG's course, which talks about YOLO, but he doesn't go into the implementation details of anchor boxes. After having looked through the code, each anchor box is represented by ...
8
votes
2answers
4k views

When should I use 3D convolutions?

I am new to convolutional neural networks, and I am learning 3D convolution. What I could understand is that 2D convolution gives us relationships between low-level features in the X-Y dimension, ...
8
votes
2answers
9k views

How to handle rectangular images in convolutional neural networks?

Almost all the convolutional neural network architecture I have come across have a square input size of an image, like $32 \times 32$, $64 \times 64$ or $128 \times 128$. Ideally, we might not have a ...
8
votes
1answer
10k views

What is a fully convolution network?

I was surveying some literature related to Fully Convolutional Networks and came across the following phrase, A fully convolutional network is achieved by replacing the parameter-rich fully ...
8
votes
2answers
250 views

Can machine learning algorithms be used to differentiate between small differences in details between images?

I was wondering if machine learning algorithms (CNNs?) can be used/trained to differentiate between small differences in details between images (such as slight differences in shades of red or other ...
8
votes
3answers
5k views

How can 3 same size CNN layers in different ordering output different receptive field from the input layer?

Below is a quote from CS231n: Prefer a stack of small filter CONV to one large receptive field CONV layer. Suppose that you stack three 3x3 CONV layers on top of each other (with non-linearities in ...
8
votes
2answers
2k views

Using neural network to recognise patterns in matrices

I am trying to develop a neural network which can identify design features in CAD models (i.e. slots, bosses, holes, pockets, steps). The input data I intend to use for the network is a n x n matrix (...
7
votes
3answers
7k views

How to use CNN for making predictions on non-image data?

I have a dataset which I have loaded as a data frame in Python. It consists of 21392 rows (the data instances, each row is one sample) and 1972 columns (the features). The last column i.e. column 1972 ...
7
votes
4answers
2k views

Why is my test error lower than the training error?

I am trying to train a CNN regression model using the ADAM optimizer, dropout and weight decay. My test accuracy is better than training accuracy. But, as far as I know, usually, the training accuracy ...
7
votes
3answers
2k views

How to compute the derivative of the error with respect to the input of a convolutional layer when the stride is bigger than 1?

I read that to compute the derivative of the error with respect to the input of a convolution layer is the same to make of a convolution between deltas of the next layer and the weight matrix rotated ...
7
votes
3answers
487 views

CNN's vs Densely Connected NN's

In image classification we are generally told the main reason of using CNN's is that densely connected NN's cannot handle so many parameters (10 ^ 6 for a 1000 * 1000 image). My question is, is there ...
7
votes
2answers
153 views

CNNs: What happens from one neuron volume to the next?

I've gone through several descriptions of CNNs online and they all leave out a crucial part as if it were trivial. A "volume" of neurons consists of several parallel layers ("feature ...
7
votes
2answers
6k views

How do we choose the kernel size depending on the problem?

Obviously, finding suitable hyper-parameters for a neural network is a complex task and problem or domain-specific. However, there should be at least some "rules" that hold most times for the size of ...
7
votes
1answer
222 views

What is the point of using 1D and 2D convolutions with a kernel size of 1 and 1x1 respectively?

I understand the gist of what convolutional neural networks do and what they are used for, but I still wrestle a bit with how they function on a conceptual level. For example, I get that filters with ...
6
votes
3answers
2k views

Does each filter in each convolution layer create a new image? [duplicate]

Say I have a CNN with this structure: input = 1 image (say, 30x30 RGB pixels) first convolution layer = 10 5x5 convolution filters second convolution layer = 5 3x3 convolution filters one dense layer ...
6
votes
1answer
541 views

Can a vanilla neural network theoretically achieve the same performance as CNN?

I perfectly understand that CNN takes into account the local dependency of each pixel to the nearby pixels. In addition, CNNs are spatially invariant which means that they are able to detect the same ...
6
votes
1answer
260 views

What does "statistical efficiency" mean in this context?

Consider the following statement(s) from Deep Learning book (p. 333, chapter 9: Convolutional Networks) by Ian Goodfellow et al. Convolution is thus dramatically more efficient than dense matrix ...
6
votes
1answer
109 views

Is there any use of using 3D convolutions for traditional images (like cifar10, imagenet)?

I am curious if there is any advantage of using 3D convolutions on images like CIFAR-10/100 or ImageNet. I know that they are not usually used on this data set, though they could because the channel ...
6
votes
1answer
762 views

Does it make sense to apply softmax on top of relu?

While working through some example from Github I've found this network (it's for FashionMNIST but it doesn't really matter). Pytorch forward method (my query in upper case comments with regards to ...
6
votes
1answer
245 views

Are there well-established ways of mixing different inputs (e.g. image and numbers)?

I am interested in the possibility of having extra input along with the main data. For instance, a medical application that would rely mostly on an image: how could one also account for sex, age, etc.?...
6
votes
1answer
2k views

How can the convolution operation be implemented as a matrix multiplication?

How can the convolution operation used by CNNs be implemented as a matrix-vector multiplication? We often think of the convolution operation in CNNs as a kernel that slides across the input. However, ...
6
votes
1answer
806 views

How does visual cortex share convolution weight

TL;DR If we buy into the idea visual cortex functions like a convolutional neural network, then there's a problem makes me scratch my head: how does brain force weight sharing as in convolutional ...
6
votes
1answer
199 views

Is one big network faster than several small ones?

The basis of my question is that a CNN that does great on MNIST is far smaller than a CNN that does great on ImageNet. Clearly, as the number of potential target classes increases, along with image ...
6
votes
1answer
292 views

Can a purely policy convolution neural network based game learn to play better than its opponents?

This question has come from my experiment of building a cnn based tic-tac-toe game that I'm using as a beginner machine learning project. The game works purely on policy networks, more specifically - ...
6
votes
0answers
96 views

How do neural network topologies affect GPU/TPU acceleration?

I was thinking about different neural network topologies for some applications. However, I am not sure how this would affect the efficiency of hardware acceleration using GPU/TPU/some other chip. If, ...
6
votes
1answer
284 views

Which neural networks are suitable for visual place recognition?

I am doing a project on visual place recognition in changing environments. The CNN used here is mostly AlexNet, and a feature vector is constructed from layer 3. Does anyone know of similar work ...
6
votes
1answer
475 views

How should the racing agent take into account the velocity of the vehicle, given the images with a speedometer?

I'm developing a game AI, which tries to master racing simulations. I already trained a CNN (AlexNet) on in-game footage of me playing the game and the pressed keys as the target. As the CNN is only ...
5
votes
2answers
428 views

Can CNNs be made robust to tricks where small changes cause misclassification?

I while ago I read that you can make subtle changes to an image that will ensure a good CNN will horribly misclassify the image. I believe the changes must exploit details of the CNN that will be used ...
5
votes
2answers
145 views

How do I improve accuracy and know when to stop training?

I am training a modified VGG-16 to classify crowd density (empty, low, moderate, high). 2 dropout layers were added at the end on the network each one after one of the last 2 FC layers. network ...

1
2 3 4 5
20