Skip to main content
31 votes

Rationale behind most published works in medical imaging trying to reduce false positives

TL;DR: diseases are rare, so the absolute number of false positives is a lot more than that of false negatives. Let's assume that our system has the same false positive and false negative rate of 1% (...
Dragon's user avatar
  • 411
23 votes
Accepted

Convolutional neural network overfitting. Dropout not helping

Ok, so after a lot of experimentation I have managed to get some results/insights. In the first place, everything being equal, smaller batches in the training set help a lot in order to increase the ...
Juan Antonio Gomez Moriano's user avatar
18 votes
Accepted

Image resizing and padding for CNN

This question on stackoverflow might help you. To sum up, some deep learning researchers think that padding a big part of the image is not a good practice, since the neural network has to learn that ...
David Masip's user avatar
  • 6,101
18 votes

Rationale behind most published works in medical imaging trying to reduce false positives

You know the story of the boy who cried wolf, right? It's the same idea. After some classifier gives false alarms (cries wolf) so many times, the medical staff will turn it off or ignore it. "Oh, ...
Dave's user avatar
  • 3,960
17 votes
Accepted

When using Data augmentation is it ok to validate only with the original images?

You should validate only on the original images. The augmentation is there so that it can help your model generalize better, but to evaluate your model you need actual images, not transformed ones. ...
Djib2011's user avatar
  • 7,998
13 votes
Accepted

Reason for square images in deep learning

There is no requirement for specific pixel dimensions for convolutional neural networks to function normally. It is likely the values have been chosen for pragmatic reasons - such as a compromise ...
Neil Slater's user avatar
11 votes

What is the distribution of categories in imagenet training set (ILSVRC2012)

I couldn't find an URL text file for the ILSVRC2012 training set, but for complete imagenet you can download the URLs only as a text file: http://image-net.org/download I wrote the following script ...
Martin Thoma's user avatar
10 votes

Image resizing and padding for CNN

You have a few options: For Small Images: upsample through interpolation pad the image using zeros If you are unable to maintain the aspect ratio via upsampling, you can upsample and also crop the ...
Ben's user avatar
  • 2,572
10 votes

When using Data augmentation is it ok to validate only with the original images?

Ideally, data augmentation is a step in your training pipeline, which comes after splitting your data into train/validation/test sets. Otherwise, you have the same data point in both training and ...
Bruno Lubascher's user avatar
9 votes

Rationale behind most published works in medical imaging trying to reduce false positives

Summary: the question probably* isn't whether one false negative is worse than one false positive, it's probably* more like whether 500 false positives are acceptable to get down to one false negative....
cbeleites unhappy with SX's user avatar
8 votes
Accepted

What is deconvolution operation used in Fully Convolutional Neural Networks?

Upsampling layer is used to increase the resolution of the image. In segmentation, we first downsample the image to get the features and then upsample the image to generate the segments. For ...
Yash Katariya's user avatar
8 votes

How can you build a model that extracts data out from receipts?

The simplest pipeline would be to do the following: OCR Named Entity Extraction Entity Disambiguation OCR This is basically transforming your receipts into plain text. If you have scans (pictures) ...
Bruno Lubascher's user avatar
7 votes

Dimension-Hopping in Machine Learning

As far as I understand the issue is the following: In image recognition the inputs to your network could be the pixels (grayscale or only 1 and 0 for black and white). If you want to, e.g. recognize ...
D. Eggert's user avatar
7 votes
Accepted

Dimension-Hopping in Machine Learning

Welcome to DataScience.SE! I'd never heard of this problem so I looked it up. It is explained on the third slide of this presentation by Geoff Hinton: More things that make it hard to recognize ...
Emre's user avatar
  • 10.5k
7 votes
Accepted

Unbalanced training data for different classes

The problem you face is commonly called the class imbalance and has been the subject of quite a bit of research. Here's a literature review, if you're interested: He, H., & Garcia, E. A. (2008). ...
Vincent B. Lortie's user avatar
6 votes

How does the bounding box regressor work in Fast R-CNN?

A very clear and in-depth explanation is provided by the slow R-CNN paper by Author(Girshick et. al) on page 12: C. Bounding-box regression and I simply paste here for quick reading: Moreover, the ...
Anu's user avatar
  • 328
5 votes
Accepted

How is the evaluation setup for YouTube faces of FaceNet?

The objective function they use to train the CNN minimizes the squared L2 distance (i.e. the squared Euclidean distance) between two similar (positive) images and simultaneously maximizes the distance ...
hbaderts's user avatar
  • 1,114
5 votes

Ways to reconstruct shuffled pixels of a video file?

A general solution to this does not exist, even if we add some assumptions about the distribution of e.g. colours and shapes in the images or temporal coupling such as consecutive frames being similar....
mjul's user avatar
  • 307
5 votes

How can I find out what class each of the columns in the probabilities output correspond to using Keras for a multi-class classification problem?

The class_indices attribute in Keras’ flow_from_directory(directory) creates a dictionary of the classes and their index in the ...
Ryan Chase's user avatar
4 votes

How many images per class are sufficient for training a CNN

From How few training examples is too few when training a neural network? on CV: It really depends on your dataset, and network architecture. One rule of thumb I have read (2) was a few thousand ...
Franck Dernoncourt's user avatar
4 votes

Training data set for food image recognition

You could use the Food 101 dataset or UEC Food 256, both contains real-world food images.
Dani Mesejo's user avatar
  • 2,226
4 votes

Convolutional neural network overfitting. Dropout not helping

There are several possible solutions for your Problem. Use Dropout in the earlier layers (convolutional layers) too. Your network seems somehow quite big for such an "easy" task; try to reduce it. ...
Andreas Look's user avatar
4 votes

Convolutional neural network overfitting. Dropout not helping

I suggest you analyze the learning plots of your validation accuracy as Neil Slater suggested. Then, if the validation accuracy drops try to reduce the size of your network (seems too deep), add ...
HatemB's user avatar
  • 326
4 votes
Accepted

Ways to reconstruct shuffled pixels of a video file?

This is a fascinating combinatorial problem. I would featuring each pixel using its full temporal trajectory, then embed them in a grid using the k nearest neighbors. The real goal is to maximize the ...
Emre's user avatar
  • 10.5k
4 votes

How does YOLO algorithm detect objects if the grid size is way smaller than the object in the test image?

Each grid predictor in YOLO should only have a high score that an object is within it, if it detects the centre of the bounding rectangle is inside itself. So a grid point that contains only the wing ...
Neil Slater's user avatar
4 votes

How does the bounding box regressor work in Fast R-CNN?

The paper cited does not mention linear regression at all. What it does is using a neural network to predict continuous variables, and refers to that as regression. The regression that is defined (...
David Masip's user avatar
  • 6,101
4 votes

How to check two images(one is original image and other one captured by mobile) similarity using deep learning?

There are a few approaches to this off the top of my head: 1. Trivial case You can first check if the images are identical by simply using numpy's functions, either ...
n1k31t4's user avatar
  • 14.9k
4 votes

Rationale behind most published works in medical imaging trying to reduce false positives

From a personal perspective, rather than a data science experience, a false positive has a higher impact on the patient's quality of live than a false negative (at least in most applications of ...
Elmy's user avatar
  • 141
4 votes
Accepted

What is the origin of YOLO/darknet coordinates

(0, 0) is top left. Here an a helpful blog that goes through all the features in the output vector. This is common in image processing. There are a few reasons ...
n1k31t4's user avatar
  • 14.9k
3 votes

Machine Learning Based Algorithm for Image Inpainting?

If I were you I would use deep learning. You can use an autoencoder format for this. Essentially you would feed the image in, then each layer yields a smaller output. Then you would feed the output in ...
Default picture's user avatar

Only top scored, non community-wiki answers of a minimum length are eligible