[Feature Request] Make image feature extractor a parameter for CombinedExtractor

### 🚀 Feature

NatureCNN is hard coded for CombinedExtractor.

https://github.com/DLR-RM/stable-baselines3/blob/master/stable_baselines3/common/torch_layers.py#L258

### Motivation
Implementing different networks requires rewriting or patching CombinedExtractor, it's a lot of code.

### Pitch

I would be useful to have it as a default rather than hardcoded.

### Checklist

- [x] I have checked that there is no similar [issue](https://github.com/DLR-RM/stable-baselines3/issues) in the repo (**required**)

Start in stable_baselines3/common/torch_layers.py around the referenced CombinedExtractor implementation and inspect how NatureCNN is selected. Determine the parameter and default behavior needed to choose a different image feature extractor, then verify that the existing NatureCNN behavior remains unchanged and that a custom extractor can be used.

Written by the indexing model from the issue text.