We are pleased to announce the Wildbook software release v.2021-09-01, which is reflected in the Flukebook.org platform for cetacean photo ID.
Wildbook DOI: https://doi.org/10.5281/zenodo.5394255 Related WBIA DOI: https://doi.org/10.5281/zenodo.4784646 For an overview of Flukebook's multi-species, multi-modal machine learning for photo ID, please see: https://docs.wildme.org/docs/researchers/ia_pipeline Community support for Wildbook is provided at: https://community.wildme.org RELEASE NOTES FOR AUGUST AND JULY 2021 CHANGES - Sighting/Occurrence search now sorts users alphabetically. - Improved Sighting/Occurence comment import behavior from bulk import. - Individual matching results correctly displays images that have been corrected with an orientation network (theta rotation). - Added additional metadata reports for bulk imports. - Added fluke matching support for gray whales. BUG FIXES - WB-1772: Corrected thumbnail orientation, which impacted bounding boxes on encounters. - WB-1741: Collaborations were causing database connections to leak, impacting memory. - WB-1730: Sequential naming for marked individuals made more robust for large catalogs. - WB-1727: Multiple attributes can be edited on an individual and encounter. - WB-1536: Button UX improved when managing the identity metadata on an encounter. - WB-1456: Merged names displaying as appropriate. - WB-605: Search results highlight in red when you need to request access to the object. WILDBOOK IA (WBIA) MACHINE LEARNING CHANGES The following updates were made to WBIA in July and August: - SAGE-20: New CurvRank V2 and PIE V2 support for gray whales added to WBIA and deployed in Flukebook - SAGE-347: Fix GPU configuration error to enable bounding box orientation (theta rotation). - SAGE-320: Accept URL and detail fields on detection API calls. - SAGE-319: Added detection status to REST API callback response. Jason Holmberg (he/him/his) Executive Director, Wild Me <https://wildme.org/> A.I. and humans combating extinction together.
_______________________________________________ MARMAM mailing list [email protected] https://lists.uvic.ca/mailman/listinfo/marmam
