Project Description

Start the activity

Objectives

Create an app to automatically identify photos of kiwi and stoats

Activity overview

This activity is divided into three parts. In the first part, the students learn about the impact of invasive predators on NZ wildlife, the need for wildlife monitoring, the concept of remote cameras, and the need for automatic recognition (~10 mins). In the second part, students create an algorithm to automatically identify animals from remote camera photos (30 mins). In the last part, the students discuss the implications of automatic photo classification for ecology and conservation, the importance of data management, and potential algorithm biases (10 mins).

Audience

NZ secondary school students (Years 11 – 12)

Curriculum Areas

Technology/Digital Technologies:

  • Use advanced programming techniques to develop
    a computer program (NZQA-AS91896)
  • Demonstrate understanding of a computer science concept (NZQA-AS91898)

Science/Science -Core:

  • Conduct a scientific experiment with guidance (NZQA-AS8096)
  • Investigate the biological impact of an event on a New Zealand ecosystem (NZQA-AS90951)

Learning outcomes

Wildlife monitoring:

  • Sampling of animal populations
  • New Zealand biodiversity
  • Data management

Train computers to recognise pictures:

  • Phases of machine learning project
  • Collecting training data
  • Interpreting results of ML
  • Common ML problems

Prerequisites

Experience with Scratch is advantageous (tutorial here)

The group leader needs to create an account for each student

Equipment needed

  • Projector
  • One computer per student
  • Access to internet