v1.0.0

[1.0.0] - 2024 - 10 - 25

Added

• Inference body-tracking components:
  • Inference data processing from Magic XRoom
  • Emotion classification inference using model checkpoint (the checkpoint trained model is not using handcrafted
    features from the training tools)
• Fusion Layer support for multi-modalities of XRoom dataset (bio-measurements and body-tracking modalities)
• Unit tests for Fusion Layer component
• Issue template for bug report and feature request.
• Docker containers registry

Changed

• Update mock writer to work with two modalities: shimmer and vr (body-tracking)
• Fusion Layer implementation for uni/multimodal: now take into account the data window (frequency * seq_len)
• Configuration file structure to encompass multimodality capability
• Body Tracking Classification now uses Pytorch instead of Tensorflow library.
• Many refactors

Removed

• Support for non-streaming inference for BM and body tracking modalities (XRoom dataset)
• ED_Evaluation component. Inference tools are tools for supporting real-world deployment of tools, so a component for
  evaluation is not part of the scope of these tools.

Full Changelog: v0.6.0...v1.0.0