concatenation

[arminak6]

Hi, congrats on the fantastic work! These results are amazing.
To address the issue of RAM usage with large datasets, is there a technique available to concatenate two outputs in case of overlap?
something like this:

`
output1 = inference(pairs1, model, device, batch_size=batch_size)
scene1 = global_aligner(output1, device=device, mode=GlobalAlignerMode.PointCloudOptimizer)

output2 = inference(pairs2, model, device, batch_size=batch_size)
scene2 = global_aligner(output2, device=device, mode=GlobalAlignerMode.PointCloudOptimizer)
`

Then concatenate scene1 and scene2 together?

Thanks again for the wonderful work!