Processes for Event-based data loading and pre-processing

[gkarray]

Issue Number:

Objective of pull request: This PR adds Processes for Event-based data loading and pre-processing.

Pull request checklist

Your PR fulfills the following requirements:

• [ ] Issue created that explains the change and why it's needed
• [x] Tests are part of the PR (for bug fixes / features)
• [ ] Docs reviewed and added / updated if needed (for bug fixes / features)
• [ ] PR conforms to Coding Conventions
• [x] PR applys BSD 3-clause or LGPL2.1+ Licenses to all code files
• [ ] Lint (flakeheaven lint src/lava tests/) and (bandit -r src/lava/.) pass locally
• [ ] Build tests (pytest) passes locally

Pull request type

Please check your PR type:

• [ ] Bugfix
• [x] Feature
• [ ] Code style update (formatting, renaming)
• [ ] Refactoring (no functional changes, no api changes)
• [ ] Build related changes
• [ ] Documentation changes
• [ ] Other (please describe):

What is the current behavior?

• No standard Processes for handling Event-based data.

What is the new behavior?

Standard Processes for handling Event-based data (loading and pre-processing) would become available.

• Currently supported Event IO Processes:
  • AedatStream : Streams event data loaded from aedat4 files.
  • DvStream : Streams event data from the DV software for INIvation cameras, through a TCP connection
• Currently supported Event pre-processor Processes:
  • BinaryToUnaryPolarity : Converts binary polarity events (0 for negative events, 1 for positive events) to unary polarity events (1 for both negative and positive events)
  • EventsToFrame : Transforms sparsely represented list of Events into a densely represented frame of Events. Two modes of operation supported:
    • 2D Output (shape (W, H, 1)) : In this case, input is assumed to be unary. Output is of shape (W, H, 1) where 0 means no Event and 1 means Event (positive or negative) in the unique channel. Negative and positive polarities are both encoded as 1.
    • 3D Output (shape (W, H, 2)) : In this case, input is assumed to be binary. Output is of shape (W, H, 2) where 0 means no Event and 1 means negative Event in the first channel and 0 means no Event and 1 means positive Event in the second channel.
• Other Process that is more general but could be used as an Event pre-processor:
  • MaxPooling: Applies the max-pooling operation on incoming data with shape (W, H, C) where C is not necessarily 1 or 2.

Does this introduce a breaking change?

• [ ] Yes
• [x] No

Supplemental information