Refactor the example and avoid hardcoded image names .

By refactoring the example, User should be able to add their own images and see the "image segmentation" happen dynamically. 

See [here](https://github.com/pytorch-labs/executorch-examples/blob/main/dl3/android/DeepLabV3Demo/app/src/main/java/org/pytorch/executorchexamples/dl3/MainActivity.java#L74-L81)