Adding Drunk effect to images of a dataset

Hi I have been working on Deep Learning recently and I have a data set which contains images of people driving inside a car. I want to add drunk class to this data set as well but the data set doesn't cover this behaviour. Is there any way to artificially add this behaviour to the images (like adding noise, rotation, intensity change) with for example matlab image processing? Can anyone suggest a method?

Answer 1

[KSSV]

Read about ImgeAugmentationDatastore.

https://in.mathworks.com/help/deeplearning/ref/augmentedimagedatastore.html

[Seyed Mousavikia]

Exactly what feture your are suggesting? I think this Augmentation only supports rotaion and resizing but what I need is way beyond that... I want to add drunk behaviour to my dataset... Can you please make your answer more specific and forgive me because I am new to this subject...

Answer 2

[yanqi liu]

yes，sir，may be use the image list to cover this behaviour，and make it to batch process，such as 3D input