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
1
vote
1answer
5 views

Difference between grayscaled and binary mnist dataset

When making a cnn you could use the classic mnist dataset containing grayscale images. I am considering transforming them to simple binary images instead, the questions is should i? It will be much ...
0
votes
0answers
16 views

How do I solve a “TypeError: __array__() takes 1 positional argument but 2 were given” Keras error?

I am trying to build a multi-input CNN using Keras/Tensorflow. I have 5000 'smile' training inputs which are 1D arrays (shape = (100,)). These inputs have a maximum length of 100. I have 5000 'protein'...
1
vote
0answers
13 views

Can Darknet be Integrated with Tensorflow and tfx?

Tensorflow and Darknet are deep learning frameworks that work and are configured. Can Darknet be Integrated differently? Is there any framework or way to integrate Darknet within a tfx pipeline, for ...
2
votes
1answer
14 views

Training a YOLO-style object detector

tl;dr I'm trying to train a small CNN (two conv layers and two connected layers) to find humans in the COCO dataset. Is my network big enough, and if so, roughly how many epochs of training will it ...
0
votes
0answers
12 views

Val loss is not reducing despite val accuracy is increasing [closed]

I am working on image classification system for which I am using Inception v3 modle. Getting above results after training is it normal. If you can suggest any research paper related to this topic
0
votes
0answers
10 views

ValueError: Invalid axis: ListWrapper([64]) [closed]

I was trying to implement the resnet50 model from scratch and this is my attempt. Defining the resnet block : ...
0
votes
0answers
8 views

SMOTE for Image regression?

Can you use something like SMOTE for an image regression task, where the target value is very skewed and imbalanced? I already tried using classic augmentation techniques like flipping, cropping, etc. ...
0
votes
0answers
10 views

Why does model CNN with high accuracy but give low prediction on handwriting input images?

I'm trying to train the CNN model with the MNIST dataset expand with my own images handwriting, so I have merged them together, ...
1
vote
1answer
27 views

Filter size in CNNs and how they relate to overfitting/underfitting

Would a smaller filter size (e.g. 3x3) potentially be more prone to overfitting than a larger filter size (e.g. 10x10) in a CNN. I know it's all dependent on the specific dataset at hand, but I'm just ...
0
votes
0answers
26 views

Is my CNN model overfitting or underfitting?

I would like to be sure of whether the model is overfitting or undercutting. Being new to this, is there any specific point to identify when to stop the training process. Any help in this regard would ...
1
vote
0answers
10 views

Choosing kernel size of cnn for time series data with multiple seasonalities

I try to solve a standard time series forecasting problem using convolutional neural networks. The data has multiple seasonalities and so I wonder if a kernel size should reflect this fact e.g. for a ...
2
votes
2answers
28 views

Less parameters - in general within ResNets

My question is about the parameters of the ResNet. Why does the network tend to have fewer parameters than the VGG? This would be the case if I got the paper and the summary from Yannic Kilcher ...
0
votes
0answers
9 views

Integrate classification models into web application C#

I created two image classificatoin models using Python (CNN) to identify car parts, and now I want to integrate them into the website of the company, they have a web application which uses C#. Does ...
0
votes
1answer
22 views

How to use efficient net as feature extractor for meta/Few shot learning in PyTorch

I am working on few shot learning and I wanted to use efficient-net as backbone feature extractor. Most of the model use Resnet as feature extractor. For example I can use below line of code and it ...
0
votes
0answers
9 views

Convolutional Neural Network CNN pixels support the last layer

I'm studying for a computer vision module and I'm on the deep learning topic, in one past paper we have the following question: Given that a convolutional neural network has five convolution layers (...
1
vote
1answer
33 views

What should the output sizing be for a class that returns multiple image arrays for a dataloader

I have a custom image class (mainly borrowed from examples) that takes in an image of size 3x244x244 for use in a VGG model and returns augmented versions (rotations of 90,180,270 and an Hflip). ...
1
vote
0answers
13 views

Working on an image classification project (microscopic images) , have some doubts [closed]

Currently, I am working on an image classification project. The data set contains very high resolution images taken via an electron microscope. Hence, I have few and limited instances. I have done EDA ...
0
votes
1answer
30 views

Why trainable parameters are not considered right?

I have tested the "ResNet" block and it works fine, but when I call it in the model class, it somehow it does not work properly? Is it related to the model definition?
0
votes
1answer
14 views

Suitable instance counting CNN for training on polygonal masks

I have a medical dataset labeled with polygonal masks (rather than rectangle boxes). It works well for pixel annotation with UNet to generate masks of healthy vs damaged skin. Now I need to do ...
3
votes
1answer
48 views

Why does adding data augmentation decrease training accuracy a tiny bit?

Before data augmentation, my model clearly overfits and hits a 100% training accuracy and a 52% validation accuracy. When only adding data augmentation with Keras, as a regularization technique, it ...
0
votes
1answer
20 views

In U-Net, is there a non-linearity (relu) in up-convolution layer?

I am doing semantic segmentatio using U-Net. I was wondering whether to include 'relu' activation or not in the up-convolution layer? ...
1
vote
0answers
8 views

Applying GradCam to video classification models

In the original paper, it says that GradCam visualization can be applied to any convolution based model. The problem is stated for convolutions that process images. In my case, I am classifying videos ...
1
vote
1answer
18 views

Pytorch: Starting with a high loss value, but the loss converged at the end. I dont know if the model could start with a loss > 100. Help!

I have been trying to attempt plant disease detection using transfer learning methods. I chose ResNet50 first. I also performed a baseline model which is a CNN model. In resnet50, I used cross entropy ...
1
vote
0answers
9 views

DCGAN - advise on why the training is not working

Objective Seeking for suggestions and advice why the DCGAN training is not working. Task Train DCGAN to learn to generate CIFAR10-like images. Each CIFAR10 image has the shape (32,32,3) where (32x32) ...
0
votes
1answer
12 views

CNN Design for Counting on Simple Images

This is the first CNN I'm designing following college examples and assignments. I'm working on a CNN that I'd like to use to classify images by the number of shapes on them. My basic problem is that I ...
1
vote
0answers
19 views

Improve CNN model for image classification

I'm using transfer learning with VGG16 for image classification. I have 6 classes each one with more than 20k images, I'm trying to improve my accuracy but after many tests I still don't have good ...
1
vote
0answers
16 views

How to interpret the confusion matrix and compare the result of features extraction with LBP and Haralick

I'm begginer in deep learning so I tried to execute a code of liveness face detection from github in this link :https://github.com/imironica/liveness , so when I tried to run features extraction with ...
1
vote
0answers
32 views

Simple implementation of 2D convolution neural network (CNN)

I am looking for a good reference implementation (code) of 2D CNN (convolution neural network) using simple math operations instead of framework's high level tensor operation. This is for inference ...
1
vote
0answers
11 views

Incompatible shape error when trying to fit my model

I am using keras to make a CNN model. I have a dataset of images which consists of training and validation tfrecord files, the images in the tfrecords are grayscale. I load in the my 2 datasets by ...
0
votes
1answer
21 views

Is it possible that MLP has better accuracy than CNN?

I am working on the epilepsy classification system which consumes EEG signals and in the result says if withing the certain period is a seizure or not. I take an advantage of Keras API for the sake of ...
1
vote
1answer
32 views

Image Classification task with unevenly sized images

Generally speaking, what's the best way to approach an image classification task with unevenly sized images (e.g. example 1 has size 300x240, example 2 has size 240x224 etc.)? Ideally I would like the ...
0
votes
0answers
16 views

How to make a prediction with an existing model in keras?

I have a pretrained model that i need to load and make a prediction on two images with. The model is a CNN that takes as input 150x150 RGB images and predicts whether it's a cat or a dog. I've sorted ...
0
votes
1answer
82 views

How to perform Multi-Label Image Classification with EfficientNet

Problem My goal is to perform multi-label image classification with EfficientNet. It should take a picture as input and e.g. tell the user that it sees a person AND a dog on the picture, meaning the ...
1
vote
0answers
18 views

Why YOLO algorithm predicts B boxes for each grid cell S?

In yolo each grid cell predicts multiple bounding boxes lets say in YOLOv1 it predicts B=2, what is the advantage as it only predicts class probabilities only once for each grid cell. If that so why ...
0
votes
1answer
52 views

How to implement random cropping during training?

I'm developing a U-net like model which segments the damaged tissue of the brain between two time-points in Multiple Sclerosis patients. The model is given the baseline and follow-up images as x and ...
0
votes
1answer
15 views

Criteria for saving best model during training neural network?

I am doing 4-class semantic segmentation with U-net using generalised dice loss as loss function. General approach to save best model during training is to monitor validation loss at each epoch and ...
1
vote
0answers
21 views

Understanding model's learning curves

I'm trying to train a Lane Detection CNN called PINet on a proprietary dataset. Below are some of the important configuration values: Batch size: 6 Optimizer: Adam Learning rate: High of 1e-4 and Low ...
3
votes
1answer
67 views

Basic doubt regarding “training” of a YOLO model

So I have just recently started exploring machine learning, and for a project I was required to train the YOLO v5 model. I first tried it on the coco128 dataset:https://www.kaggle.com/ultralytics/...
0
votes
0answers
29 views

CNN Color Invariance

For the task of detecting an object of any color, if we train our CNN only with images that only contain that object in one or two colors, will the accuracy of our model's predictions be affected for ...
1
vote
1answer
27 views

Count repeating “objects” in a picture

This is my first data-science project and I would love to get some guidance to know how to get started. My problem is the following: I want to count objects that are in a picture. This picture has a ...
0
votes
0answers
10 views

DQL for detecting next move in games

I am trying to understand using DQL for playing board games and how we can do function approximation of the q-learning Bellman equation in order to detect the best next move , if anyone can give the ...
0
votes
0answers
12 views

Tensorflow process killed

I need to train 5 models each for 10 times using tensorflow and keras for a homework. 2 of the models are multilayer perceptron models and remaining 3 of them are CNN models. I am training using my ...
1
vote
1answer
18 views

Keras: apply multiple filters to each feature map in CNN

I am new to Keras, and I want to do the following: take a 2D image, and apply four 2D convolution kernels to it, giving four 2D feature maps. I could accomplish this. But then I want to apply two ...
0
votes
2answers
56 views

Jupyter, Python: the kernel appears to have died while training a model on a big amount of data

I am training my model on almost 200 000 images, i'm using Jupyter and now after 3 days of training ( i used 800 epochs and batch-size = 600) I have this " the kernel appears to have died. It ...
0
votes
2answers
20 views

Zero Padding in Convolutional neural network

We use Convolutional neural network because it by design learns features that generalize over spatial location , so when using conv operation it reduce image size and that what we hope to have so we ...
0
votes
0answers
13 views

Improving a CNNs accuracy - help & advice

So I have created a CNN for image classification, and I train and test it with two datasets. One contains 9,339 images and the other 9,100 images. The first model which I designed gave an accuracy of ...
0
votes
0answers
19 views

Sklearn compute_class_weight function leads to inaccurate result

I am trying to fit CNN to imbalanced data with 3 classes and therefore I am using compute_class_weight function from SKLearn. The code is presented below. ...
0
votes
1answer
20 views

CNN model accuracy

I have trained my CNN model on CIFAR 10 and I got val_accuracy of 87% which is not a low value but when it comes to detection of pictures my model detected most of the pictures wrong. anyone knows why ...
0
votes
1answer
38 views

CNN model low accuracy

I have 1299 images in 4 classes (374/269/284/372). I want to use the VGG19 model, add a dense layer at the top and fine-tune it with my images. As I only have 1299 images, I also want to use data ...

1
2 3 4 5
24