Database: Mesh2PPM 1.0.0: Training (2022-2024)

Subtitle: Randomly varied local BezierPPM parameters to generate the BezierPPM instances used to train Tool Mesh2PPM 1.0.0
Cite as: Pausch, F., Perfler, F. ., Holighaus, N., and Majdak, P. (2025). "Mesh2PPM 1.0.0: Training", The SONICOM Ecosystem: Database #52. URL: https://ecosystem.sonicom.eu/databases/52. Copy Citation to Clipboard

Metadata

Creators:

Publishers:

Rightsholders:

Keywords:

Relations:

Other:

  • DOI: not assigned yet
  • Uploaded by: Florian Pausch ORCID: 0000-0003-2728-3170 Email address: florian.pausch@oeaw.ac.at
  • Date (created): 2025-12-01 15:54:36 (GMT)
  • Date (updated): 2025-12-05 14:02:53 (GMT)
  • Production Year: 2022-2024
  • Resource Type: Dataset (SONICOM Ecosystem)
  • Rights: EUPL-1.2: European Union Public Licence version 1.2
  • Subject Areas: Life Science , Other SONICOM Ecosystem
  • General Description: Parameters of Tool BezierPPM 3.0 to generate synthetic pinna geometries, multi-view grey-scale pinna images and multi-view depth images with Tool PyBezierPPM 3.0. The resulting BezierPPM instances and renderings were used to train Tool Mesh2PPM 1.0.0, an AI-based estimator of pinna parameters from pinna meshes.
  • Technical Remarks: Contains BezierPPM parameters as CSV files. Use function export_ppm() in create_data.py, which is part of Tool Mesh2PPM 1.0.0, to generate meshes (STL files) of the corresponding BezierPPM instances, and to render these meshes as multi-view grey-scale images (PNG files) and depth images (OpenEXR files). For rendering, the camera views were either set to nominal (camera_jitter=False) or randomly deviate from nominal (camera_jitter=True).
  • Data Source: Computer simulated

Comments

No comments found.


Uploaded by: Florian Pausch
Created: 2025-12-01 15:54:36
Updated: 2025-12-05 14:02:53