image Post Your Answer


image

Neural network for pre-processing image data


Is neural network the best option for pre-processing image data? Also, how do I build predictive models for images?

All Answers (3 Answers In All)

By Sallehahsim Answered 5 years ago

To build predictive models for images, you have to first develop training images. These images must be then processed to make a new prediction. The system used to process the images would be neural networks. The neurons in the neural network will take 1 pixel of the image and apply simple computation such as activation function to the piece of the image. The output is then fed to extraneural layers till the process is completed and a prediction model is built.


By Shanti Answered 5 years ago

You can definitely use neural networks to pre-process the images. The factors to be considered to perform this process are image size, number of images, number of channels, aspect ratio, image scaling, dimensionality reduction, and data augmentation.


By Nkumardo Answered 5 years ago

You can use neural networks for image processing. But it has few limitations. Debugging and interpreting results is made difficult and fully connected architecture would be sufficient to process huge image data.


Your Answer


View Related Questions