Questions tagged [convolutional-neural-network]

A convolutional neural network is a form of neural network with an additional convolutional layer, typically used in image & audio analysis. The convolutional layer is essentially a filtering stage defined by the kernel which is used. For example, a convolutional layer could have a kernel which extracts edges from an image towards the goal of learning which objects are in a scene.

Filter by
Sorted by
Tagged with
0
votes
0answers
13 views

Representing multi-channel input signals with a single signal

I am working on an EEG signal classification problem. My dataset consists of EEG signals stored as 19X30000 NumPy arrays. Each row represents a single channel. For now, I am converting each of the ...
0
votes
1answer
66 views

Testing accuracy very low, while training and validation accuracy ~ 85%

I have a training dataset of 10000 pictures and a test dataset of 15000 pictures. There are 23 types of birds. First of all, I imported the necessary ...
0
votes
1answer
31 views

How to calculate convolution for 2nd conv Layer in CNN, Do we need to average across all feature maps?

I understand that for the first layer (assuming we have a grayscale image) we calculate the convolution of 3*3 receptive field as a weighted sum of receptive weights with pixels $ x1 · w1 + x2 · w2 + ...
0
votes
0answers
14 views

Adding layer to a trained CNN to process higher resolution images. Tried 2 schemes, 1 works fine, 1 fails completely

I'm working with images coming from a sensor, for which 1 pixel corresponds to 2 mm in the real world. I've built and trained a CNN that does semantic segmentation of the image (128x128 pixels) and it ...
3
votes
0answers
31 views

Using softmax for multilabel classification (as per Facebook paper)

I came across this paper by some Facebook researchers where they found that using a softmax and CE loss function during training led to improved results over sigmoid + BCE. They do this by changing ...
0
votes
0answers
40 views

Pytorch: how to pass the hidden state between the samples in LSTM?

I am trying to boost the performance of a object detection task with sequential information, using ConvLSTM. A typical ConvLSTM model takes a 5D tensor with shape ...
0
votes
0answers
19 views

Best 5-layer pretrained CNN model

I am doing a visualization project on convolutional neural nets to aid learning and need a simple to display but complex enough pretrained CNN model so I can visualize feature maps for each layer. I ...
0
votes
0answers
24 views

Getting constant accuracies for training and validation sets despite their losses are changing during CNN training?

As the title clearly describes the issue I've been experiencing during the training of my CNN model, the accuracies of training and validation sets are constant despite the losses of them are changing....
0
votes
1answer
37 views

Training loss = 0, training accuracy =1, validation and test around 85%

I have created different CNNs for doing image classification. The dataset is this: https://www.kaggle.com/crowww/a-large-scale-fish-dataset There are 9 classes, and each class contains 1000 images of ...
0
votes
1answer
29 views

Max Pooling in first Layer of CNN

I am seeing, in all the notebooks that I found, that Max Pooling is never used in the first layer of a CNN. Why this? Is it a convention among data scientist to do not use max pooling in the first ...
0
votes
0answers
14 views

How to deal with a small dataset for image classification using CNN?

I have a dataset consisting of characters(lowercase and uppercase) and numbers, totalling about 62 classes. The data I have are about 45 images per class and no test data. The data is a subset of the ...
2
votes
1answer
112 views

Improve Convolutional Autoencoder

I just built a Convolutional Autoencoder to try to reconstruct a time series with shape (4000, 10, 30). This is the code, I used a batch size of 32, but I think it ...
1
vote
1answer
18 views

3 images as one input in CNN (U-Net) [closed]

I have been advised by my supervisor that if my U-Net segmentation network has RGB images at the input then I could use the channels for different images - median filter for R, normalization for G, ...
0
votes
0answers
29 views

Parallel programming in Python

I have the next code that I am trying to run in parallel: ...
0
votes
1answer
35 views

How to use data generator for regression keras?

I am using the Keras data generator to load data from a directory. I am basically dealing with a regression i.e there is a numerical value for each of my images in the range ...
1
vote
0answers
20 views

Compare distance between embeddings in different dimensions

I am working on a problem with CNNs. After the convolutional layers, comes a "flatten". One could interpret that as a representation of the input image in some high-dimensional continuous ...
0
votes
0answers
13 views

Using Wasserstein loss function for image-to-image-regression

The context I have a 3D array (representing a grayscale 3D image) and want to turn this into another 3D array of the same size. In this output array the value of each pixel is a number that measures ...
0
votes
0answers
26 views

Which applications can not be handled by very Deep CNN models?

I wanted to know what challenges very deep models can face even if the accuracy is good. Would they be not suitable for any application given that my model is very very deep? I wanted to know if ...
0
votes
0answers
6 views

Multi-object detection within single image

Given an image with multiple objects within it, I would like to train a CNN to output vector of labels corresponding to the presence/absence of objects within the image. I would like to know whether ...
1
vote
0answers
58 views

Calculate the Convolutional Autoencoder sizes - Conv1D

I'm approaching the Conv1D for the first time and I do not understand how to calculate the parameters in each layer. I have an input of (3000, 10, 30), but I ...
1
vote
0answers
139 views

Understanding scipy.signal.convolve2d full convolution and backpropagation between convolutional layers

I'm learning about convolutional neural networks. The convolution operation in order to extract features that is described in literature and posts used for this is quite intuitive and easy to ...
1
vote
0answers
27 views

Create sequence for a Conv1D layer

Im studying the following tutorial on the Keras website and I'm trying to understand how to create a sequence for a Conv1D layer. This is their method: ...
1
vote
0answers
14 views

How to modify a Convolutional Neural Network architecture built for a univariate time series to multivariate time series?

I have built a CNN (in combination with a LSTM cell) that takes 1D time series-like data as an input and performs classification. I am obtaining a good performance, but the complete data has actually ...
1
vote
0answers
13 views

Convnet with peculiar loss function not learning!

Im using this loss function: ...
2
votes
0answers
25 views

Methods to visualize the filters in the later layers of a CNN?

I've extracted the weights from the filters of a pretrained model (AlexNet). I wish to represent these weights visually, this works fine for the first layer as there is only 3 input channels so I can ...
3
votes
1answer
52 views

Is there an appropriate use of adjusting class weights for a balanced dataset?

I ask this because I am currently working with a CNN model built for diagnosis of pneumonia. Originally, I followed a notebook on kaggle to build the model and thereby learn what each bit of code is ...
0
votes
0answers
19 views

Validation Accuracy not going beyond 60% for image classification with 5 species of snake

My dataset has about 17000 images belonging to 5 classes. I am using 16000+ images for training(about 3k/class) & 500 for validation(100/class). Training accuracy is very good but validation ...
0
votes
1answer
81 views

What will be the input_shape of tf.keras.layers.Conv3D be for these inputs

I have many videos, and each video is made up of 37 images (there are 37 frames in the whole video). And the dimension of each image is (100, 100, 3).... So the ...
0
votes
0answers
9 views

Use CNN for document digitalization

I'm using Faster RCNN to digitalize a document. It is able to recognize zones of the document (name, name of enterprise, name of director, etc) but cannot classify them accurately. So, my problem is ...
1
vote
0answers
28 views

How does time needed for training differ between different batch sizes?

I've constructed a CNN in Python using Numpy, which is trained with mini batch gradient descent for MNIST digit classification. When training with a batch size of 1, the time needed for 5 epochs is ...
2
votes
1answer
151 views

TypeError: Expected int32, got None of type 'NoneType' instead

I want my model batch size to be a dynamic shape, and I've assigned none as batch size, but that's causing an error. Here, in the first line, I specified batch size as None: ...
3
votes
0answers
11 views

Running DenseNet from cmd line and jupyter but vastly Different loss and accuracy [closed]

I’m running a DenseNet121 from Pytorch with the same exact code, same exact hyper parameters and same exact image sizes, once from a jupyter notebook and once using the command line via a python ...
1
vote
0answers
18 views

Should a filter learned within a residual block be different form its vanilla CNN counterpart?

I have a very basic CNN using Conv2d with multiple layers and activations, each layer $\ell$ has parameters $w_\ell$ inducing a mapping $f_\ell(x,w_\ell)$. Now I decide to introduce skip connections ...
0
votes
0answers
13 views

Multidimensional Output from Radar Imagery and Climate Data

I am trying to predict what my rainfall field will look like at a future timestep using: Radar imagery of rainfall fields at previous timesteps: A set of 2D matrices where each element in each matrix ...
1
vote
0answers
20 views

How Does EAST detector implementation with VGG16 look? How many outputs does it have?

I was reading the Efficient and Accurate Scene Text Detector paper and saw the author reference VGG-16 as a possible stem "feature extractor" network. In the paper they say: In our ...
0
votes
1answer
29 views

Siamese vs matching network for correct image category matching

I have to find the closest match between my image and bunch of already collected images of different classes in the folder. Whic meta-learning approach should I select. I am thinking about the Siamese ...
1
vote
0answers
22 views

Trade-off between number of channels and size of convolutional filters

As far as I understand, the common practice in the modern CNN architectures is to use a smaller convolutional filters, but deeper networks with more channels. One of the reason behind this is that one ...
0
votes
0answers
39 views

Questions about adding metadata to a CNN using keras

I have a convolutional neural network and would like to include some metadata. My metadata is in a multiple csv files that correspond to each class and it contains a bunch of geometric properties (...
-1
votes
2answers
21 views

Optimisation of neural networks

Do neural networks get optimized by trial and error, by data scientists, or is there some way of optimizing values through accurate mathematical equations?
1
vote
0answers
11 views

How to prove Separable Convolution layer is theoretically identical to traditional Convolution?

I have seen the saying that Separable Convolution layer is theoretically identical to traditional Convolution for so many times, but yet no one has pointed out where the proof is. God, I have google ...
0
votes
1answer
244 views

Conv1D layer input and output

Consider the following code for Conv1D layer ...
1
vote
0answers
25 views

Padding in Convolution Formula

Why is it that the formula for each element in a convolution between an image $I$ and a $k \times k$ sized kernel $K$ is $$ (I*K)_{ij}=\sum_{m=0}^{k-1}\sum_{n=0}^{k-1}I_{(i-m),(j-n)}K_{mn}=\sum_{m=0}^{...
0
votes
0answers
16 views

Understanding the convolution formula

According to several sources this formula, or the center originated version of it, is used to calculate an element of a convolution between an image $I$ and a kernel $K$ of size $k \times k$: $$ (I*K)...
4
votes
0answers
325 views

1D CNN Variational Autoencoder Conv1D Size

I am trying to create a 1D variational autoencoder to take in a 931x1 vector as input, but I have been having trouble with two things: Getting the output size of 931, since maxpooling and upsampling ...
-1
votes
1answer
70 views

Issues in plotting Images using Keras

I am trying to visualize Skin Cancer Images using Keras. I have imported the images in my notebook and have created batch datasets using Keras.image_dataset_from_directory. The code is as follows: <...
0
votes
1answer
265 views

Autoencoder implementation using ImageDataGenerator

I'm using the concept demonstrated in this paper. Their training data consists of "GOOD" images and "BAD" images. They train the AE using "BAD" images (X) to make it ...
0
votes
0answers
24 views

DCGAN: why does my generator has less loss then my discriminator?

I have constructed a DCGAN (deep convolutional generative adversarial network) inspired by this github repository. It is written in a more low level Tensorflow code that I tried transforming into ...
0
votes
0answers
24 views

Use convolutional variational autoencoders for time series prediction

I want to use convolutional variational autoencoders for time series prediction. For example, here is the dimension of my data. ...
3
votes
1answer
30 views

Does a Convolutional Layer in a Neural Network learn the correlation between its input signals via its kernel?

I am interested in the theory behing what a convolutional neural network learns with its convolutional operations. I think it learns (useful) kernels which measure the correlation between its input ...

1
2
3 4 5
12