How do I get one overall prediction, where each data point has many pictures?

Data Science Asked on January 2, 2022

My task is not a simple image -> category. I have between 5 and 10 images of an object, and I must classify it. The problem is that the category isn’t "visible" in each picture, so I would need many images -> category.

For instance, let’s say I’m trying to classify men vs. women. I have a number of pictures taken from random angles. I need to determine, based on the aggregated score of these pictures, which category the bunch belongs to.

All the pictures contain a little bit of unique information.

classification cnn deep learning machine learning neural network

One Answer

You are describing a problem of supervised learning with multiple inputs. That is not an uncommon task and you can find many tutorials about multiple inputs for neural networks out there. Using Tensorflow, I personally recommend Keras Functional API for this task, since it gives you more control on the layers while keeping the high-level simplicity.

Answered by Henrique Nader on January 2, 2022

How do I get one overall prediction, where each data point has many pictures?

One Answer

Add your own answers!

Related Questions

How will my CNN results be affected by large discrepancies between the number of samples in some of the classes?

Train NN for balanced accuracy

Horizontal Bar Char Issues in Tableau with Excel Spreadsheet Beginner Visualization Question

Should Principal components be normalized before applying K means on them?

Naive Bayes and Support Vector Machine (NBSVM) Classification

feature importance with CNN

How to find whether a dataset is blanced or imbalanced?

What is the defining Set in NLP

Understanding declared parameters in my Conv2d layer of my convolutional neural network

How to add Error Bar to Matplotlib line plot?

Why some papers write false positive rate per case instead of percentage rate?

how to find parameters used in decision tree algorithm

What is the difference between “Adding more LSTM layers” or “Adding more units on existence layers”?

Add variables/features to already trained tensorflow model when new variables come in. Possible?

Time series stationarize vs normalization

How to use a dataset with only one category of data

How do we standardize arrays with NaN?

Data augmentation for recommendation systems

Omnibus and R square improvements for OLS model

Why is it important to have stable activations in the forward pass?

Ask a Question