Cellular neural network

Why do image processing tasks take longer as images get bigger?

Imagine you're sorting photos on your phone. The more photos you have, the longer it takes to find a specific one. Now, think about processing an image to recognize objects; the larger the image, the more time it takes.

As the image gets bigger, there's more to check for patterns and features. It's like searching for a word in a longer sentence; you have to read more letters.

Example

If sorting 10 photos takes 2 minutes, sorting 100 photos might take 20 minutes.

Remember this

The computational cost is O(n²) because every pixel's calculation depends on its neighbors, creating a grid-like pattern of dependencies.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

Types of artificial neural networks

Why can CNNs spot patterns better than RNNs?

to use a CNN: for data with spatial structure like images or time series

Why can't we just feed all data into one big neural net?

batch size affects generalization: larger batches find sharper minima

Larger batch sizes lead to sharper minima, enhancing generalization by providing more accurate gradient estimates

gradient accumulation simulates larger batch sizes without more memory

Can you train a machine like you do with a computer?

2022 in science

Why do Transformers sometimes seem to 'ignore' irrelevant parts of the input?

Convolutional neural network

Can a neural network learn too well?

Swipe through more Machine Learning concepts

Open Pocket Polymath