We are pleased to announce the Wildbook software release v.2021-05-31,
which is reflected in the Flukebook.org platform for cetacean photo ID.

 Wildbook DOI: 10.5281/zenodo.4895243
 Wildbook Image Analysis (WBIA): 10.5281/zenodo.4784646

For an overview of Flukebook's multi-species, multi-modal machine learning for photo ID, please see:


Community support for Wildbook is provided at:


Release notes for May 2021

Flukebook Updates
 
- Added new locations in the Atlantic Ocean and Hawaii.

- Improved algorithms for gray whale computer vision have been deployed for performance testing

Bug Fixes

 - WB-1671 Resolve submitter email issue

 - WB-1634 Images no longer swapping width and height with annotation location.

- WB-1630 Conflict between projects and submissions has been resolved.

- WB-1626 Corrected “My data” restriction in match filtering.

 - WB-1625 Corrected project filter in iaResults to include all project matches.

 - WB-1601 Google maps renders GPS and ignores LocationID correctly.

 - WB-1529 Flukebook projects display more than the IndoCet Opportunistic project.


**Wildbook-IA Release Notes (Machine Learning)**

The following updates were made to WBIA last month.

SAGE-40 Completed the AI extension for the MAPS project  (general purpose, multi-species dorsal fin matching). Added bonus: works great for white shark individual ID too!

SAGE-293 Two new publications at CVPR workshop “CV 4 Animals” on 1) the continual ID curation problem and 2) CurvRank V2.

SAGE-304 New WBIA API for the background job engine allows job IDs to be specified instead of generating random UUIDs.

SAGE-310 Migrate WBIA codebase in GitHub to use the new “main” branch convention.


Bug Fixes

SAGE-292 Fix EXIF orientation parsing with OpenCV.

SAGE-309 Fix automated testing on CI for GitHub.



Jason Holmberg (he/him/his)

Executive Director, Wild Me

A.I. and humans combating extinction together.