Category

BRAIN UP YOUR DEVICE

Applications (apps) are not only valuable software tools for mobile devices. They also play an increasing role in software built in other devices and machinery to fulfil helpful tasks where mobile phones fall behind. These can be as different as stationed electronic devices for monitoring purposes or software built in medical instruments. Often these applications include neural networks help machinery work autonomously. Also sometimes mobile applications serve the purpose of stearing or interacting with attached devices directly.

Sunbird images has experience in developing apps for devices other than mobile phones such as the DBU founded project to build an automatic pigeon song detector called “Birdtector“. On the birdfair in August 2017 Sunbird proposed the idea of developing an optical device to automatically identify animal species. Many fruitful discussions followed and April 2019 the Swarovski “dG Finder” was released, with Software mainly delivered by Cornell University and Sunbird Images. Swarovski created the monocular field glass, capable of identifying birds and mammals in conjunction with accompanying apps with the “dG Mammals” app built by Sunbird Images.

Project

SWAROVSKI OPTIK  dG Mammals

Sunbird Images was assigned by SWAROVSKI OPTIK Austria to develop the app for automatic image recognition of mammals, which was to be used with their dG device on both, iOS and Android mobile devices in the form of smartphones and tablets.

Development

CHALLENGES

The first major challenge was the selection of species included in the app. The goal was to enable the user to automatically determine European mammals and to provide him with informative and interesting facts. The app contains information about plus automatic recognition for such diverse groups as rodents, European game, and even marine mammals. Here, the aim was to include all mammals that can typically be observed with long-range optics and that are desired by a wide range of users. For this purpose, Sunbird Images conducted an intensive user analysis to enable an abundant selection of species while making sure that the “must have” mammals are definitely included in the app.  

The heart of the app, however, is of course the AI to identify mammals. This AI was trained on the basis of the latest technology by Israeli machine learning experts and using hundreds of thousands of animal images. A special feature at Sunbird Images is that the training images do not simply come from ready-made databases, which often contain wrongly identified species, but have been carefully curated by Sunbird Images staff. This, and the inclusion of the recognition of the corresponding habitat of an animal, results in a very high accuracy.   

An additional feature implemented by Sunbird Images is that the app receives data directly from the SWAROVSKI OPTIK dG device, which means it must be connected to the device at all times. This is not only important for the immediate automatic identification of the observed animals, but also for the live stream. This innovative feature transfers the image the dG operator is currently observing to the phone or tablet with the device. Especially for tour and nature guides, this is a fantastic opportunity to show their customers animals in the wild on a tablet screen without having to explain on which branch or stone the animal currently sits or behind which leaf it is hiding. A very true example of “Digital goes Nature!“   

UX/UI

In order to combine the various requirements with a user-friendly and intuitive framework, Sunbird Images created a UI that is simple to use, yet offers a user experience that reflects SWAROVSKI OPTIK’s high quality standards and their premium impression. Sunbird Images’ designers and user experience experts worked together, supporting responsive design for the optimal display on different devices embracing SWAROVSKI OPTIK corporate identity. The iOS and Android OS frameworks for the app both have the same user interface structure, but were programmed and implemented independently for the respective operating systems.