E heke e Heka vs Aotearoa Species Classifier

The Unlocking Curious Minds fund supported Scion to develop ‘E heke e Heka!’ a new interactive learning app designed to empower rangatahi (young people) and community with knowledge about myrtle rust. With a focus on accessibility and engagement, the app combines current science information with creative and culturally relevant content, promoting Te Reo Māori and Reo Rua bilingual language options for diverse audiences ages 12 years and above. In the myrtle rust virtual learning ngahere, the user journeys through stages, gaining knowledge on different aspects of science related to Myrtaceae – myrtle plants, the origin of myrtle rust, infection, identification, environmental factors, and outlook for the future. The app steps the user through environmental science content in a digestible way, incorporating digital illustrations and photo reference visuals, interactive elements, games, and augmented reality activities that connects the app user with the science knowledge presented. FEATURES Myrtle Rust science script in Te Reo Māori and reo rua / bilingual / English languages Synchronised narration of Te Reo Māori and scientific names Swipe to read and /or Touch to hear Export and share learning summary pages and images Interactive games to assist embedding learning outcomes This app was developed by Scion and produced by world leading cultural services agency Kiwa Digital. Any Scion existing and created Intellectual Property rights reserved. For more information / assistance: support@kiwadigital.com

An app for classifying species found in Aotearoa New Zealand. It is based on an artificial neural network model trained on 14991 species, including plants and animals, that can be found around New Zealand. It is possible to take images by camera or select images from local storage. The model runs locally, so no network connection is required to make classifications. The imagery used for training the artificial neural network is a snapshot of all research-grade observations from Aotearoa New Zealand available via the crowd-sourcing website iNaturalist. Improvements to the model will be possible primarily by retraining the model on a more recent snapshot. Users are encouraged to upload their observations to iNaturalist, particularly for species that do not currently have many observations made in Aotearoa New Zealand. To facilitate upload, it may be best to take photos outside the app, so that a photo can be uploaded to iNaturalist via a suitable app if appropriate. Note that the classifier does not know at all about species that were not represented in the snapshot of observations used for training. In particular, invasive species that are new to Aotearoa New Zealand or those for which iNaturalist did not have any observations from Aotearoa New Zealand when the model was trained will not be known to the model.