Questions tagged [convolutional-neural-networks]
For questions about convolutional neural networks, also known as CNN or ConvNet.
889
questions
0
votes
0answers
9 views
How to normalize images before training?
I have seen people normalize images by just dividing 255. But why? Why not use mean normalization or Z-score Normalization? I also came across this StackOverflow topic while searching but the answers ...
1
vote
0answers
11 views
Why identity mapping is so hard for deeper neural network as suggested by Resnet paper?
In resnet paper they said that a deeper network should not produce more error than its shallow counterpart since it can learn the identity map for the extra added layer. But empirical result shown ...
0
votes
0answers
13 views
How to set weights from TimeDistributed layer on normal layer? [closed]
I want to get some weights from the big model which uses TimeDistributed layer and set these weights in some smaller convolutional model.
As far as I know if you use TimeDistributed the weights are ...
1
vote
0answers
22 views
Setting up a deep learning architecture for multi-dimensional data
The input data is thousands, millions of 4x1000 matrices. Each row consists of 3 small natural numbers (1000 combinations) and a corresponding real number between 0 and 1.
The output is a 1x1000 ...
2
votes
2answers
50 views
How do we know that the neurons of an artificial neural network start by learning small features?
I'd like to ask you how do we know that neural networks start by learning small, basic features or "parts" of the data and then use them to build up more complex features as we go through ...
1
vote
1answer
18 views
Why are CNN binary classifier output probability distributions often skewed?
I've been working on a lot of simple resnet18 binary classifiers lately and I've started to notice that the probability distributions are often skewed one way or the other. This figure shows one such ...
0
votes
1answer
10 views
What are the state-of-the-art Person-Detektion / Human-Segmentation?
I would like to use a deep learning approach to detect people in videos. I have found some freely accessible implementations like Human Segementation with Pytorch or BodyPix / DeepLab / Pixellib with ...
2
votes
0answers
21 views
What are the metrics to be used for unsupervised monocular depth estimation in computer vision?
I am currently replicating the results of this paper. In this paper they have not mentioned how they are evaluating the results as no ground truth is available for comparison. Same goes for other ...
3
votes
1answer
57 views
GAN Generator Output w/ Periodic Noise
I am training a Semi-Supervised GAN, using multivariate time-series with window of shape (180*80) with the generator and discriminator architecture below. My data is scaled using Robust Scaler, so I ...
1
vote
1answer
39 views
Is it a good idea to use different width and height of the kernel in a CNN?
I always see that the width and height of the kernel are the same. But is it a good idea to use different numbers?
Recently I tried to use GoogLeNet (which expects images to be 224x224) on my images (...
4
votes
2answers
362 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 ...
0
votes
1answer
28 views
Converting age and sex variables to a 64-unit dense layer
I am studying a preprint for my own learning (https://www.medrxiv.org/content/medrxiv/early/2020/04/27/2020.04.23.20067967.full.pdf) and I am befuddled by the following detail of the neural network ...
2
votes
1answer
267 views
Is there any difference between ConvNet and CNN?
ConvNet stands for Convolutional Networks and CNN stands for Convolutional Neural Networks.
Is there any difference between both?
If yes, then what is it?
If no, is there any reason behind using ...
1
vote
0answers
26 views
Error in MobileNet V1 Architecture?
From the architecture table of the first MobileNet paper, a depthwise convolution with stride 2 and an input of 7x7x1024 is followed by a pointwise convolution with the same input dimensions, 7x7x1024....
3
votes
0answers
33 views
Best Machine Learning Model for “Predicted” Image Generation
I am currently working on undergraduate research to determine hotspots for hand-surface contact. Ideally, I would like to give the model a depth image as input:
Example of synthetic depth image
and ...
2
votes
0answers
33 views
CNN leaf segmentation throught classification of edges how to improve
I am trying to design a CNN that can do pixel wise segmentation of edges leaves in dense foliage agriculture images. Such as these:
On the basis of this article https://arxiv.org/pdf/1904.03124.pdf, ...
3
votes
1answer
60 views
How do you calculate KL divergence on a three-dimensional space for a Variational Autoencoder?
I'm trying to implement a variational auto-encoder (as seen in Section 3.1 here: https://arxiv.org/pdf/2004.06271.pdf).
It differs from a traditional VAE because it encodes its input images to three-...
1
vote
0answers
16 views
Dealing with bias in multi-channel auto encoders
The problem
I have a multi-channel 1D signal I want to auto-encode.
I am unable to resonstruct the input when the number of channels increases.
Code
I am using a convolutional encoder, and a ...
1
vote
0answers
24 views
Does the order of data augmentation and normalization matter?
What is the preferred order of data augmentation and normalization? Is it the former followed by the latter?
0
votes
0answers
26 views
Improving Mask RCNN by arbitrary scaling head input
Currently, I am looking at how Mask R-CNN works. It has a backbone, RPN, heads, etc. The backbone is used for creating the feature maps, which are then passed to the RPN to create proposals. Those ...
0
votes
0answers
24 views
Are all activation functions in a layer same? [duplicate]
I understand that for you can have multiple activation functions in different layers. CNN's usually have Relu followed by softmax for the classification. But what stops us in having multiple ...
0
votes
1answer
36 views
Denoising Images When Training a Classification Model
Suppose you have a binary outcome variable and have some training data (10,000 images in jpg format). Also you have a test set of say 11,000 images. If we want to train a classification model and want ...
3
votes
0answers
36 views
Is the GAN architecture better suited for medical image denoising than the CNN?
I'm considering using GANs for medical image denoising, based on previous literature, like this and this. My input to the GAN would be a high-noise image and my ideal output would be a low-noise, high-...
0
votes
0answers
36 views
Can I use a CNN for template matching, so that there is robustness, as the background of the target image is not that good?
I have to extract part of a source image, then I have to check if it is similar or almost similar to any of the 10 target images, so that I can do further processing on that one specific target image, ...
0
votes
0answers
27 views
Yolo from scratch dataset and output
Hi I coded a YOLO model from scratch and just came to realise that my dataset does not fit the models output. This is what I mean: The model outputs a ...
2
votes
0answers
44 views
Estimating dimensions to reduce input image size to in CNNs
Considering input images to a CNN that have a large dimension (e.g. 256X256), what are some possible methods to estimate the exact dimensions (e.g. 16X16 or 32X32) to which it can be condensed in the ...
2
votes
1answer
64 views
How to incorporate a symmetry constraint in the loss function to train a CNN?
I have a task of extremely sparse binary segmentation, i.e. the segmentation mask contains either 0 or 1, and there are ~95% zeros and only ~5% ones. I use the focal loss to address the sparseness (...
1
vote
0answers
17 views
Underfitting a single batch: Can't cause autoencoder to overfit multi-sample batches of 1d data. How to debug?
TL;DR
I am unable to overfit batches with multiple samples using autoencoder.
Fully connected decoder seems to handle more samples per batch than conv decoder, but then also fails when number of ...
2
votes
1answer
62 views
What do the variables in the cross-correlation formula mean?
I understand what cross-correlation does given a kernel and an input image, but the formula confuses me a little. Given here in Goodfellow's Deep Learning (page 329), I can't quite understand what $m$ ...
1
vote
0answers
39 views
Understanding Batch Normalization for CNNs
I am trying to understand how batch normalization (BN) works in CNNs. Suppose I have a feature map tensor $T$ of shape $(N, C, H, W)$
where $N$ is the mini-batch size,
$C$ is the number of channels, ...
1
vote
1answer
47 views
What are acting as weights in a convolution neural network?
Looking at some old notes I took on CNN's and I wrote down that the weights in a CNN are acting like filters in a CNN but to be honest I don't really know what the weights are acting as in a CNN and ...
1
vote
1answer
58 views
Predicting continous value with CNN (prediction of fruit maturity)
I want to train some IA algorithm to be able to evaluate the maturity of a fruit (say, measured in numbers of days before rotten) based on an image of the fruit.
My first instinct is to go with ...
1
vote
1answer
30 views
Keras 1D CNN always predicts the same result even if accuracy is high on training set
The validation accuracy of my 1D CNN is stuck on 0.5 and that's because I'm always getting the same prediction out of a balanced data set. At the same time my training accuracy keeps increasing and ...
1
vote
1answer
40 views
What happens if there is no activation function in some layers of a neural network?
What if I don't apply an activation function on some layers in a neural network. How will it affect the model?
Take for instance the following code snippet:
...
0
votes
0answers
15 views
How to get the prediction probability of random sample image from multiclass classification model?
I am performing classification using AlexNet as transfer learning(simply say performing classification using CNN) for five types of class on 18000 images. These 18000 images are divided into Train, ...
2
votes
1answer
40 views
Is it possible to train one part of the network with a particular learning rate and the other part with a different one?
I have a combined network consisting of two parts: one is for images and the other is for numerical data. Each sample is matched with a numerical case by an ID. For this combined network, a ...
3
votes
1answer
44 views
Can a convolutional network predict states for a RL Agent
During the course of training a DQN agent, all visited states are stored in a replay buffer. Therefore would it be practically possible for a CNN, given a reasonable amount of data, to predict the ...
1
vote
2answers
70 views
How to extract parameters from a text using AI/NLP
lets say I have three texts:
"make a heading that says hello word"
"make a heading of hello world"
"create heading consist of hello world"
How can I fetch those groups ...
2
votes
0answers
25 views
Graph Neural Networks: Quesitons about different GCN Architectures
This might be moreof a question about nested function classes:
For k class node classification in a graph with n nodes, and d feature vector.
I want to compare
Architecture I: the GCN model of Kipf/ ...
1
vote
1answer
46 views
How do you pass the image from one convolutional layer to another in a CNN?
I am currently trying to write a CNN from scratch, but I don't understand how to feed the information from a max-pooling layer to the next convolutional layer. Specifically, I don't know what to do ...
1
vote
0answers
20 views
In the DeepView paper, do they use the same FCN for all depth slices AND all views?
I'm trying to replicate a paper from Google on view synthesis/lightfields from 2019: DeepView: View Synthesis with Learned Gradient Descent and this is the PDF.
Basically the input to the neural ...
0
votes
2answers
24 views
Semantic segmentation CNN outputs all zeroes
I'm using MATLAB 2019, Linux, and UNet (a CNN specifically designed for semantic segmentation). I'm training the network to classify all pixels in an image as either cell or background to get ...
3
votes
1answer
57 views
What is the use of the regular convolutional layer in expansion path of U-Net?
I was going through the paper on U-Net. U-net consists of a contracting path followed by an expanding path. Both the paths use a regular convolutional layer. I understand the use of convolutional ...
1
vote
0answers
25 views
How do gradients are flown back into the Siamese network when branching is done?
I am curious about the working of a Siamese network. So, let us suppose I am using a triplet loss for my network and I have instantiated single CNN 3 times and there are 3 inputs to the network. So, ...
1
vote
0answers
36 views
What are the pros and cons of 3D CNN and 2D CNN combined with optical flow for action recognition?
For action recognition or similar tasks, one can either use 3D CNN or combine 2D CNN with optical flow. See this paper for details.
Can someone tell the pros/cons of each, in terms of accuracy, cost ...
1
vote
0answers
21 views
Is there any research work that shows that we should explicitly mark the word boundaries for 1D CNNs?
I'm doing character embedding for NLP tasks using one-dimensional convolutional neural networks (see Chiu and Nichols (2016) for the motivation). I haven't found any empirical evidence of whether or ...
0
votes
1answer
40 views
How is the latent vector transforming to a feature map in DCGAN (Generator structure)?
I'm working on the code trying to generate new images using DCGAN model. The structure of my code is from the PyTorch tutorial here.
I'm a bit confused trying to find and understand how the latent ...
1
vote
0answers
14 views
Considerations when doing image classification where the object is not the subject
I've come across two types of image classification tasks
cat/dog classification the whole picture is either a cat or a dog. Simple.
this image contains a cat classification. There's a whole chaotic ...
1
vote
1answer
39 views
Output volume proof for convolutional neural network
As I've been dabbling into the sliding window concept, I stumbled on a question that asked me to find the number of windows needed on a 1D image of $W$ size, knowing the window size $K$ and the stride ...
1
vote
2answers
73 views
Is there anything that ensures that convolutional filters end up different from one another?
I found this question very interesting, and this is a follow up on it.
Presumably, we'd want all the filters to converge towards some complementary set, where each filter fills as large a niche as ...
