Questions tagged [cnn]

Convolutional Neural Networks (CNN, also called ConvNets) are a tool used for classification tasks and image recognition. The name giving first step is the extraction of features from the input data.

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

Why does BatchNorm on FC layers increase my error?

I am training a deep CNN for multivariate regression, with three fully connected layers on top of the convolutions. I am using Sigmoid activation for FC layers. When adding BatchNormalization (BN) I ...
0
votes
0answers
28 views

Keras using weights from one CNN on another with same architecture except Input size

Related to TrackNet, a CNN for tracking tennis balls on TV tennis matches, the Arxiv paper mentions it is scalable, ie. the input can be any number of frames concatenated rather than the three they ...
0
votes
0answers
8 views

How to correctly use depthwise convolutional layers

I am trying to speed up my CNN by replacing all convolutional layers with depthwise convolutional layers, which can require only as much as $10$% of the operations ...
0
votes
0answers
12 views

Input 0 is incompatible with layer conv1d_40: expected ndim=3, found ndim=2

i am working on computer vision using deep learning. my training data contains (x,128) shape. i am passing the same to conv1d layer but facing issues below is my model ...
2
votes
1answer
34 views

What are the exact differences between Deep Learning, Deep Neural Networks, Artificial Neural Networks and further terms?

After having read some theory I am getting a bit confused about the following terms: Deep Learning Deep Neural Network Artificial Neural Network Feedforward Neural Network So, what seems clear to me ...
0
votes
0answers
27 views

Join Multiple Tensor from a CNN features extractor

I have a mixed neural network. The first part is a CNN that extrapolate features from an image; I used: ...
2
votes
2answers
30 views

In CNN, why do we increase the number of filters in deeper Convolution layers for complex images?

I have been doing this online course Introduction to TensorFlow for AI, ML and DL. Here in one part, they were showing a CNN model for classifying human and horses. In this model, the first ...
1
vote
1answer
24 views

Is there a way to get y_pred values from saved keras model?

I have a saved Keras model that saved in .h5 file, as you know there are a y_pred and y_act ...
2
votes
2answers
39 views

predicted bounding boxes that stretch beyond grid cell (Andrew NG CNN course)?

I was following Pr. Andrew Ng course on Course about Convolutional neural network and I have a doubt regarding one of the points he mentions in the ...
0
votes
0answers
11 views

Hardware needed for training VGG16 from scratch

i'm working on a classification problem in which i need to classify with the highest accuracy possible 38 classes. Here is a link to the dataset: PlantVillage dataset I want to get the highest ...
0
votes
1answer
15 views

Extract segment from document scan

I need to extract some "valuable" information from document scan. For example, document's number, incoming date, organizations, persons, etc. Example document: I'm trying to extract highlighted ...
1
vote
1answer
16 views

Classification of non-stationary acoustic signals

I'm currently working on a task of classification of short non-stationary audio signals with length of 1024 elements and sampled at 120 kHz. I was wondering if there are any special techniques or ...
0
votes
1answer
20 views

Non-Real Time Data Augmentation for CNN Classification. What are the drawbacks?

When people talk about and use data augmentation, are they mostly referring to real-time data augmentation? In the case of image classification, that would involve augmenting the data right before ...
1
vote
1answer
15 views

Training a CNN to convert ellipses into circles

My current project has to do with modeling the effects of blurring/convolution of objects in various imaging processes. Right now, I am starting off with a preliminary, artificial model. I am using ...
0
votes
1answer
23 views

Does adding of many FC layers during re-training increase the model size ? Are there any ways to optimize the size of model?

I am re-training a pretrained model VGG16. In the last layers, im using two FC layers of size 2048 each, with dropout=0.5. When i saved the model, The size of the model was found to be 2 GB (which ...
0
votes
1answer
12 views

CNN image to image translation: multiple image inputs to one image output

I am interested in training a CNN to take in inputs where each input is a set of low-resolution images and each ground truth is a single high-resolution image. The ground truth high-resolution image ...
1
vote
0answers
12 views

Keras: Merged/Concatenated model perform worst than separate models memes recognition

I have a dataset of memes, and I'm trying to predict if a certain meme is sexist or not, using image and text together. Right now I have two models, a ...
1
vote
2answers
26 views

What is the dimension of the filters if the input image has only one channel?

I have a grayscale image with dimension HxWx1 (one channel). To build a CNN using the grayscale image as an input image, what is the dimension of the filters? I read from some websites, it says that ...
0
votes
2answers
24 views

What can be the best dropout value and the FC layer for good accurate predictions?

I am retraining the pre-trained model VGG16 in the last FC layers. I used the below function . what can be the best combination of FC layers and dropout values for the best predictions. ?
0
votes
1answer
16 views

Pytorch convolution input reshaping

I am new to CNNs, and I'm trying to follow along with a Pytorch DCGAN tutorial by reimplementing it in Keras. Clearly there are some differences in the frameworks, but in particular I am struggling to ...
0
votes
0answers
7 views

model training with center and softmax loss

i want to train a cnn with about 5000 classes and inception resnet v2 using center loss but at first it decreases and then it starts increasing. Furthermore the evaluation loss is very large.I trained ...
0
votes
1answer
13 views

How to use decode_predictions() for non-Imagenet models..?

I know that decode_predictions() works for only imagenet dataset(1000 classes) for the models like VGG16 etc. But condiser my scenario. My Scenario: I used vgg16 pretrained model, and added my own ...
1
vote
0answers
19 views

Combination of two CNN models outputs

I have an image data set that I want to classify into 3 classes: non-disease, disease-type 1, and disease-type2. I trained two separate CNN's models, one is classifying the 3 classes, and the second ...
0
votes
1answer
13 views

How to find the class name of a new image from the pre-trained model

I would just like to get the class names of the predictions. I can get the class names on the images that I trained the model. But if I predict an image (say which is not trained but already belongs ...
0
votes
0answers
31 views

My model is overfitting though i'v been using regularization techniques

I've been training an Xception model to recognize the disease of a plant from its leafs. So far i reached a training accuracy of 91% but the test accuracy is around 73%. So obviously my model is ...
0
votes
1answer
24 views

while re-training a pre-trained model, I'm facing this issue RuntimeError: You must compile your model before using it

model summary: RuntimeError: You must compile your model before using it. It says that the model needs to be compiled. But as far i know, if i compile a model, all the previous trained data will be ...
0
votes
0answers
12 views

Algorithms for change detection in images

I'm wondering why people use complex CNNs like in Unsupervised Change Detection in Satellite Images Using Convolutional Neural Networks in order to find the ...
0
votes
2answers
23 views

How to train or approach the image datasets with different resolution in Deep Learning

Image classification: I am having a data set of image collection more than 10k but even though all are the same image but taken in different sizes (pixels into pixels) some are in square and some are ...
0
votes
0answers
9 views

DC GAN to W GAN conversion

I have implemented DC GAN network and I know that changing the loss function , we get a W-GAN network, but I wonder how to code the wasserstein loss function and integrate it with my code below: Here ...
0
votes
1answer
16 views

Dimension change of convolutional layer after applying the feature map

In this lecture I don't understand how the output has 1 layer after applying the feature map to the 3 layer input.
0
votes
0answers
10 views

Input tensor for CNN should be of shape (n_rows, n_timeSteps, n_features) or (n_rows, n_features, n_timeSteps)?

I have a regression task in which I want to predict the value y based on X values (with 4 features X1, X2, X3 and X4). For demonstration purposes I generate some random data: ...
1
vote
0answers
17 views

Visualizing general adversarial network

I am working on a DC GAN model using my own data set. How can I visualize (see output) of the GAN generator to see how my network is working (replicating the training data)? Here is my code: ...
2
votes
5answers
56 views

How to get high accuracy in CNNs?

ML and Data Science world. I am a newbie to CNNs, but do possess a basic understanding of ML and Neural Networks. I wanted to create my own CNN that works on the Cats and Dogs Dataset. I ...
0
votes
0answers
15 views

How to Reduce Overfitting of Deeplearning models on NLP tasks in unbalanced datasets?

I have a binary classification problem, where the number of examples belonging to Class 0 is 20% on average. And the rest 80% of examples fall into Class ...
0
votes
0answers
11 views

How to use WaveNet or dilated CNN with multivariate time series

Typical WaveNet and dilated convolution network are used to work with simple time series with one observable, that is the shape of input data is ...
2
votes
0answers
50 views

How to add a CNN layer on top of BERT?

I am just playing with bert (Bidirectional Encoder Representation from Transformer) Research Paper Suppose I want to add any other model or layers like Convolutional Neural Network layers (CNN), Non ...
0
votes
1answer
17 views

How to approach the PlantVillage dataset?

I'm working on the PlantVillage dataset and i want to predict the type of the disease from the image of a leaf. The dataset is labeled in pairs (Type of the plant,Healthy/name of the disease). I'm ...
0
votes
0answers
18 views

Training a CNN on a probability map

Lets say you want to detect something of a picture made with the bird perspective (ie with a drone). Can you create a dataset with the pictures and the self pre generated probability maps as labels, ...
0
votes
0answers
9 views

Plotting the value change in regression CNN

I have trained my CNN on thousands of images with different hyperparameters, now I want to show the difference on the output for each individual image. I mean ...
0
votes
0answers
62 views

Transfer learning on yolo using keras

I am working on a project that uses object detection. I have logo images that need to be detected in a video. I am doing this in keras. I followed this blog to convert the yolo weights to a keras ...
1
vote
1answer
73 views

Using LeakyRelu as activation function in CNN and best alpha for it

Since if we do not declare the activation function, the default will be set as linear for Conv2D layer. Is it true to write: <...
0
votes
0answers
8 views

Unsupervised CNN - STN

I would like to implement unsupervised CNN to make affine registration. My input : 64x64x1, binary image of square which are warped, translations and rotations. I use normalize Cross Correlation for ...
0
votes
1answer
11 views

CNNs - Hyperparameter tuning with different training sizes of the same data set

I would like to compare how much the classification performance (test accuracy) of CNNs changes depending on the size of the data set. For this I would like to use a data set like MNIST or Fashion ...
0
votes
0answers
17 views

The effect of removing pooling layers in the model's accuracy

I know that removing pooling layers will lead to an increase in dimensionality and subsequently, make the training to be more time-consuming. But I'm wondering if ...
0
votes
0answers
18 views

Minimizing cross entropy vs minimizing negative probabilities

The cross entropy loss can be written as $L_1 = -\sum_i\sum_c y_{ic}\log P_{ic}$, where $i$ represents images and $c$ are the classes. $y_{ic}=1$ for the correct class. Instead of L_1 I can minimize ...
0
votes
0answers
17 views

Help needed implementing Convolutional Sequence-to-Sequence Network

I am trying to build convolutional Sequence-to-Sequence network that takes inputs (satellite images) and predicts the next sequence of images. As a result, we can then predict the weather. I have ...
0
votes
1answer
21 views

Is text recognition by definition a part of image recognition?

I'm referring to more advanced text recognition systems that are using neural networks to find and extract text from images like the ones Google and Microsoft are offering on their ML platforms. If ...
0
votes
0answers
18 views

How to deal with small datasets for CNNs?

I have a dataset with 2000 images and 4 labels. I'm going to use data augmentation(brightness, flip, skew, etc) and use pre-trained weights from vgg-16. However, i'm not sure this will do the trick. ...
1
vote
0answers
15 views

Slowly decreasing validation / training cost and their abnormal values

I have a dataset of size ~100,000 of images, I'm training a CNN model on them for regression. optimizer: Adam batch_size: 64 Number of epochs: 50 When I set the ...
0
votes
1answer
32 views

Extremely slow CNN

I am trying to train a CNN with keras in R. I have a time series that is 3-dimensional, so every sample has dimensions 95 x 365 and has 80 features, which I feed in as channels. The output is only 1 ...