Test With Own Data

[jeffin07]

@roatienza Thank you for the code. while testing I've some doubts.

1. Can we try densemapnet using our own data? If so can you give us an example
2. Why do we need to download the dataset for testing? For testing, the checkpoint is enough right?

I tried to test the densemapnet using predictor.py but it gives an error of
FileNotFoundError: [Errno 2] No such file or directory: 'dataset/driving.train.disparity.1.npz

So I downloaded the sample of driving dataset but it still shows the error