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.
1,375
questions
0
votes
0
answers
9
views
Training and Validation accuracy decreases after using more data
I have a binary classification project, I use a neural network with the following architecture:
The shape of the input is 64×64×4. This input was fed to a Conv2D layer with 32(5×5) filters followed by ...
0
votes
0
answers
6
views
Why does this cGAN model perform poorly when trained on a different machine?
I was sent a cGAN model python file from a friend + the dataset he used to train this model. For him, the model trained succesfully & was able to generate very accurate images. These were his ...
- 1
0
votes
1
answer
29
views
ReLu layer in CNN (RGB Image)
I am able to get convoluted values from RGB Image lets say for each channel.
So I have red channel with values: -100,8,96,1056,-632,2,3....
Now what I do is that I ...
- 103
0
votes
1
answer
107
views
Pytorch mat1 and mat2 shapes cannot be multiplied
The error message shows
RuntimeError: mat1 and mat2 shapes cannot be multiplied (32x32768 and 512x256)
I have built the following model:
...
- 27
0
votes
0
answers
10
views
why by adding additional information as number of sequence on dataset can avoid overfitting
I am developing a regression model to analyze walking styles. The dataset I am using to build the model is from 2 different sources, let's call them dataset A and dataset B. Dataset A has a shape of (...
- 21
0
votes
1
answer
17
views
Why don't we increase the parameter from 64 to 128 in this CNN model?
I'm looking at an example lab from a coursera course titled Intro to Tensorflow. In this CNN model, they're gradually increasing the no. of filters from 16 to 32 and then 64. Why don't we increase it ...
0
votes
0
answers
17
views
why CNN the model can't predict 0
I have two datasets: force plate data and plantar pressure data. The force plate data consists of 6 data points, while the plantar pressure data consists of 90 data points. Both datasets have a ...
- 21
0
votes
1
answer
23
views
Is deep learning high initial validation accuracy a sign of problem?
I have a image classification model with 8400 images of class A and 1800 images of class B. I have used validation_split=0.2 with subsets of ...
- 111
1
vote
0
answers
20
views
Object Detection: Setting threshold values as trainable parameters?
I am building my first object detection model (Mobilenet SSD, to detect animals in images) and happy with the current test results.
When I tested it using images without bounding boxes, I noticed some ...
- 11
-1
votes
1
answer
23
views
Determining "filters" dimension after a convolution operation
I tried to calculate the "filtered" dimension and I seem to be getting it wrong.
Below there is the image I am trying to calculate the "filtered" dimension for, where you have 192 ...
- 3
0
votes
0
answers
19
views
CNN model well trained but can't predict real data
I'm developing a CNN regression model for gait analysis. It seems the model is well trained, with low val_loss and low loss. However, the model does not work well to predict real data. In this ...
- 21
0
votes
0
answers
23
views
why MSE will be high when I inverse data
I make a regression model to predict force plate using plantar pressure. I am trying to use CNN model in this case. I have 2 different datasets, dataset A (force plate data) and dataset B (plantar ...
- 21
0
votes
0
answers
17
views
How to visualise a feature map or filter in a group equivariant convolutional neural network
I'm reading an article called "Group Equivariant Convolutional Networks" by T. Cohen and M. Welling (https://arxiv.org/abs/1602.07576) and I'm having some problems understanding one of their ...
- 1
0
votes
1
answer
30
views
How to build an image generation model for interior room design?
I want to build an image generator model of interior room design. This model should be able to generate an interior image of a living room/bedroom/hall/kitchen/bathroom. I have searched about it and ...
- 51
0
votes
0
answers
14
views
Which CNN model to use for the classification(20 classes) of gemstones (diamonds, sapphire, ruby etc) based on digital photo images and huge data set?
Im trying to build CNN Model for the classification of precious stones (like diamonds, sapphire, ruby) based on digital images. So I have data set of labeled 150,000 gemstone certifications and the ...
- 1
0
votes
0
answers
16
views
Grad-CAM for CNNs with GAP layer
I'm new to deep learning, so maybe this is a silly question...
Do any adjustments need to be made for applying Grad-CAM on CNNs that use a Global Average Pooling (GAP) layer right before fully ...
0
votes
0
answers
9
views
Are there any advantages on considering images as graphs and use them on Graph Convolutional Networks?
I have seen this encoding of an image as a graph:
The set of the nodes $V$ is the set of pixels. If the image is of size $10\times10$, then we have $10\cdot10=100$ pixels.
Each node has a length 3 ...
0
votes
0
answers
20
views
Would a CNN be an appropriate model for a board game?
I want to make a reinforcement learning algorithm for a game. It's a turn-based game that takes place on a board. You would basically have 4 typical actions: attack and rush the enemy, attack and flee,...
0
votes
0
answers
16
views
What are some methods to convert time series data into images for CNN?
I am working on a project where I have specific time series data which I would like to convert to images. I have investigated various methods, such as Markov Transition Fields, Gramian Angular Fields, ...
- 3
0
votes
0
answers
17
views
Grad cam outputs for all the images are the same
I am using grad cam to see which regions of the test images are most important for the prediction of resnet50. The output I got has some errors.
Code Snippets:
<...
- 172
0
votes
0
answers
20
views
Can early stopping be used with stratified k-fold validation to help avoid overfitting in neural networks?
I am using stratified cross validation and using transfer learning for the classification. I have 4 classes. I am training my model with fold=10 and epoch=20.
code snippets:
...
- 172
0
votes
0
answers
30
views
Regularization in CNN change training accuracy but not validation accuracy
I have a question regarding regularization in convolutional neural networks.
So I'm building a CNN for image classification and I've come across something I don't understand.
Without using any ...
1
vote
1
answer
32
views
Validation acc is very high in each fold but Test acc is very low
I am trying to implement a neural network. I am using CNN model for classifying. First I split the dataset into train and test.
Code Snippet:
...
- 172
0
votes
0
answers
8
views
Shaping Data for time series ConvLSTM
I am having the same problem and I am unable to properly fit the input to the model. Can u please share some of your code snippets. mainly for input data reshaping and passing into the ConvLSTM2d. ...
- 1
-1
votes
1
answer
82
views
Manual computation of the predictions in a convolutional neural network
I am trying to manually compute the predictions of the Keras library for a convolutional neural network. However, I am struggling a lot to match my final result with the ones provided by Keras. I do ...
- 59
3
votes
2
answers
151
views
Measure distance between teeth using Machine Learning
I'm a newbie in ML and I have a problem I am stuck on.
I want to train a ML model to recognize dental diagnosis based on photos and x-rays of the patient. Specifically right now, I want to find a way ...
0
votes
1
answer
28
views
A curve val_loss and loss in keras after training a model
Can anyone help me, is my model overfitting or underfitting?
I want to make sure the model is well done before starting to deploy
Also, I use categorical cross-entropy loss
I have asked before, but I ...
- 1
-1
votes
1
answer
49
views
Curve val_loss and loss in keras after training a model
I trained a Keras model to diagnose disorders and want to make sure it is good enough to start deploying.
From the below graph, can anyone advise me as to whether my model is overfitting or ...
0
votes
0
answers
59
views
In CNN, When do we increase, or decrease, the number of filters/neurons?
Good morning,
I would like to understand how do we choose between increasing or decreasing the number of filters applied in a CNN.
My logic response to this, would be to take Autoencoder as an example ...
- 1
0
votes
0
answers
69
views
WARNING:tensorflow:Your input ran out of data; interrupting training Error
Due to VRAM capacity limitations, I cannot fit the whole training and validation data into the GPU memory. Instead of cutting some of the data out, I decided to use TF.dataset object to create the ...
0
votes
0
answers
13
views
Validation accuracy decreases when I use ImageDataGenerator instead of manually creating labels
I was getting 85 percent validation accuracy before using ImageDataGenerator, now validation accuracy for binary classification decreases even to 14 percent. and never passes 60 percent. I couldn't ...
- 1
1
vote
0
answers
92
views
problem on designing a custom_loss function
I am using CNN to solve a regression problem in a supervised manner. i have input data(X_train) and the target data(y_train). I allow the network to train and during training process in each batch of ...
- 73
0
votes
0
answers
36
views
Adding Data sequences as unique data on dataset for regression model
I want to predict a force plate using plantar pressure. The shape of the force plate data is a 15000x6 array, and the shape of the plantar pressure data is a 15000x89 array. I will use a regression ...
- 21
0
votes
0
answers
31
views
Binary Classification: My model classfies most data (95%+) as label 1
I am working with ECGs and trying to use a CNN model to perform binary classification. The goal is to classify 30s ECGs to detect a specific disease. I am using CNN and converting ECGs to images (...
- 11
1
vote
2
answers
113
views
Should a CNN generalize to arbitrary positions in the data?
I have trained a CNN on one dimensional data that is the power spectral density (PSD) of a $N$ different classes of signals ($N=4$). Each of the $N$ signals has a different spectral shape (not shown ...
0
votes
0
answers
13
views
Using features extracted from CNN and handcrafted features to perform classification
I have a question in regards to merging features extracted from CNN and handcrafted features.
I have been reading this paper https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9002895/#B33-sensors-22-02467
...
- 11
0
votes
0
answers
17
views
Best Tool for NN with Convolutional Layers
I am working on creating a NN with the following architecture:
Input layer (180 neurons)
Hidden Layer 1 (18 neurons)
Hidden Layer 2 (4 neurons)
Output Layer (1 neuron)
I am trying to figure out the ...
- 101
1
vote
0
answers
12
views
How can I test a model on longer sequences that it is trained on
I am trying to train a model on ECGs. Unfortunately, I do not have enough data for 30s ecgs but I have sufficient data for 10s ecgs. I have trained a CNN model and performs really well on the 10s data....
- 11
0
votes
1
answer
54
views
Can I have 0 loss in the validation set and still have bad accuracy?
I am starting in the world of deep neural networks and doing a series of tests with a convolutional model, I have found the following case:
The accuracy in the training set is much better (around 0.85)...
- 302
1
vote
1
answer
65
views
Random Forest Classifier is giving me an array of zeroes
I used VGG16 as feature extractor on a dataset with 9 classes and trained the Random Forest Classifier on the feature vector. I tried to make prediction on the test feature vector but the prediction ...
- 23
0
votes
1
answer
43
views
How to build model if the data dont have corelation each other's
I have 2 datasets, call them dataset A and dataset B. Then I want to predict dataset A using dataset B as input using regression model.
dataset A format:
dataset A shape(15000,1)
dataset B format:
...
- 21
0
votes
0
answers
8
views
How to select best kernel_size and max_pool_size in CNN1D
I have data with shape size 1,89. setup kernel_size = 3 and pool_size = 2 on the conv1d layer. However, the model is not able to predict the peak well. i think the problem is because the kernel_size ...
- 21
0
votes
0
answers
11
views
How the low level features are combined to form high level features in CNN? What happens to combine low level feature to form higher ones in bw layers
I want to understand basics behind cnn features formation like how high level features are formed using low level features in a CNN?
1
vote
0
answers
29
views
Why is my segmentation model not returning a heat map?
I have implemented two CNN architectures to perform segmentations on medical images: the classic UNet and a modified version called the Attention UNet. I have been training the models on roughly 50,...
- 11
1
vote
1
answer
75
views
why CNN model can't learn well the peak from data
here I have two different datasets. dataset1 is force plate data and dataset2 is plantar pressure data. dataset1 has shape (2050,2) and dataset2 has shape(2050,89). before doing the training I have ...
- 21
0
votes
2
answers
21
views
In a convolutional layer, is it standard practice to modify stride and padding to get a desired output?
I'm trying to implement the CNN described in A Framework of Hierarchical Deep Q-Network for Portfolio Management (see screenshot). In the paper, the author describes the first CNN layer as having a ...
0
votes
0
answers
15
views
Image recognition model with CNN for face gestures is really bad
I have a dataset that contains facial expressions and their label, and I am trying to make a classification model for it. Unfortunatly, I can't manage to create a good model with CNN, as the highest ...
0
votes
1
answer
26
views
Sound Anomaly Detection
What is the recommended directory structure for sound anomaly detection using Keras CNN (Unsupervised) ?
After converting the sound files into spectrograms. Code examples will be highly appreciated.
- 1
3
votes
1
answer
70
views
Solving video classification problem by taking EVA Large as backbone
I am solving a video classification problem. There are 9 classes in total. At first I took ResNet as a feature extractor, this gave me 0.74 accuracy. Then I changed ResNet to EVA (I also tried Swin), ...
- 131
0
votes
0
answers
24
views
High Recall and Low Precision for Binary CNN model
I was training a CNN model for binary classification. The training and validation accuracy seemed good. However, the precision is low and the recall is high (High false positive).
...
