Q learning algorithm in image processing using matlab.

I would like to process the images by removing noise from the blur and color correction using Q-learning simpler and more efficient because artificial intelligence and especially deep learning consists of more advanced methods on image processing. I know about neural networks but i don't know much about this algorithm. so, i am totally confused from where to start and how can i do this. Can anyone please share some link because i searched a lot from past two days and i saw most of the videos of q-learning algo is about either games or robot. I can't find any help on image processing side. Please help me, i have also the option of doing this on python.

Answer 1

[Emmanouil Tzorakoleftherakis]

Hello,

Finding an example that exactly matches what you need to do may be challenging. If you are looking for the "deep learning" version of Q-learning, it's called DQN. See some details here and some examples below.

https://www.mathworks.com/help/reinforcement-learning/ug/create-agent-using-deep-network-designer-and-train-using-image-observations.html

https://www.mathworks.com/help/reinforcement-learning/ug/train-dqn-agent-to-balance-cart-pole-system.html

If you keep finding examples of Q-learning for the application you are thinking maybe there is a reason for that. Without being an expert, Q-learning works with tabular policies, so it may fit nicely with a scenario where you want to manipulate pixels in an image (which is a table after all)

[NABIHA ANAYA]

Can i do this Q learning part on "DEEP NETWORK DESIGNER" in matlab?

Actually i have no idea about this app as i didn't work here before, but i will try....thanks

[Emmanouil Tzorakoleftherakis]

If you were to use DQN (not Q learning), you could use Deep Network Designer to design the critic network if you wanted to. Q learning works with tables so Deep Network Designer is not relevant

[NABIHA ANAYA]

So, which app is suitable for Q learning? Because i have to do Q learning and after searching i find out reinforcement learning toolbox used in this. But i haven't free trial.

[Emmanouil Tzorakoleftherakis]

There is no specific app for Q learning. I would follow the example here to see how to set up the problem