Data Science Stack Exchange
Data Science Stack Exchange

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.

1367 questions
5
votes
2 answers

How filters are made in a CNN?

I am new to Data Science and CNN. My understanding of CNN is that: An image's pixel data is convoluted over with filters which extract features like edges and their position. This creates filter maps. Then we apply max pooling which will down…
Lakshay Dulani
  • 265
  • 2
  • 6
5
votes
1 answer

Combining spatial input with a label as input for CNN using Keras

I also asked this question on Stack Overflow. However, it has not yet been answered and I think this is a more suitable platform to place it. I'm trying to implement a network set-up similar to this Google Deepmind paper. Their network set-up is as…
Boris Mulder
  • 83
  • 11
4
votes
1 answer

Generating image embedding using CNN

I have a CNN model using cifar -10 dataset. The model was built using Keras (Tensorflow). Now based on this model, I have to generate an image embedding (vector). That means - an input image comes and I have to output the embedding vector of that…
nad
  • 163
  • 1
  • 1
  • 5
3
votes
1 answer

Accuracy decrease in production after adding additional input datas

I am trying to predict TimeSeriesA by using a CNN. I create snapshot images of the timeseries and these are then labelled. With a very simple snapshot I get reasonable training and test accuracy. When I apply the model to real world in production I…
ManInMoon
  • 75
  • 6
3
votes
1 answer

How to combine multiple images features

I have a dataset that contains information of pets (Breed,color,age,some text descriptions) and images of that pet. One pet can have more images than others. I want to combine these images somehow to generate a single 1x255 feature so that I can…
Ha An Tran
  • 131
  • 1
3
votes
1 answer

How is bias added in a convolutional layer

In a typical neural network , bias is usally added like this v = activation ( w1*x1+ ... + Wb*b) However, I am not really sure how it is done in convolutional layer. My one thought is that it is added with each convoltional operation for a neuron.…
user58529
  • 31
  • 1
  • 2
3
votes
1 answer

How multi-scale CNN selects final output map

I read a few days ago about multi-scale CNN (OverFeat method) which you can access to presentation via this link. They performed CNN on different scales of an image and then combine all output maps. They said inside of that…
saeed masoomi
  • 146
  • 8
2
votes
1 answer

How does CNN doing bounding box regression and what do features and weights represent for?

I knew that, in the house price logistical regression problem, the weights and features represent the "importance" of factor or coefficients of feature variables respectively, then minimize LSR loss can get the value of coefficients, and question…
Alex Luya
  • 133
  • 5
2
votes
1 answer

What morphological operations can be used to vectorise road surface binary rasters?

We are currently developing a Deep Convolutional Neural Network to extract road surfaces from aerial orthophoto's. Our current workflow involves using an existing road centreline to mask a large enough area to determine which of the pixels are road…
Peter Wilson
  • 121
  • 2
1
vote
0 answers

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 only. Many open source code implementations out are…
Dilshod Tadjibaev
  • 111
  • 2
1
vote
0 answers

Draw a CNN architecture in Latex

Is there a way that I can draw a simple CNN architecture in LaTeX if I know that my model has: four convolutions, three blocks (convolution + BN + RELU + dropout), and then pooling after each layer? Then lastly a normal convolution Conv2d.
pochi
  • 11
  • 1
1
vote
0 answers

Logit Data for CIFAR-10

I am working on a project that requires the logit outputs (pre-softmax layer) on an accurate model for the CIFAR-10 dataset. I have been looking at Densenets since it has a softmax output layer. However, the current implementations that I've found…
Jason
  • 11
  • 1
1
vote
0 answers

CNN - Confused on the output shape of second convolutional layer

I'm attempting to write a forward pass of a CNN but I'm stuck on the second convolutional layer. From what I understand, given an image of size 28x28, a first filter of size 10x3x3, a second filter 20x3x3, and max pooling after each filter, the…
Yeti.91
  • 11
  • 2
1
vote
2 answers

How to maintain accuracy with more classes?

I'm working on a CNN model (implemented by keras) that detects landmarks from images. Every landmark (for example Acropolis, Big Ben, Eiffel Tower etc) is as a separate class and is divided into own directory. The training dataset contains 1600…
Tauno
  • 799
  • 2
  • 9
  • 9
1
vote
2 answers

Why does my model sometimes not learn well from same data?

I have a dataset of 2 classes, both containing 2K images. I have split that into 1500 images for training and 500 images for validation. This is a simple structure for testing purposes, and each image is classified depending on the colour of a…
ManInMoon
  • 75
  • 6
1
2 3