Objective : To help children identify species of birds through bird songs using efficient Deep Learning based Audio Classification models

Concept : Children can enhance their knowledge of different species of birds after iterative hearing of common bird songs and can become more aware of the local fauna around them using smart web/mobile based applications with simple but powerful user-interface backed by Artificial Intelligence and Machine Learning

Issues addressed: Children are natural born explorers filled with curiosity regarding their surrounding environment. They tend to wonder about different sounds which they come across, be it a squirrel’s squeak or a sparrow’s chirp. They need to be able to differentiate between the two and in this manner become aware about the species of birds that live in the area along with other exotic species in different parts of our Country. We need an mobile/web-based application that can help children quickly and easily identify birds, record their findings and learn more about them.

Details : Identification of species of birds within a surrounding can be an important factor for a child to have an overall understanding of fauna in their community. The app that we are building will have an inherent database of bird sounds and their corresponding species name. Using this app, a child will be able to record and upload the bird sound and using the underlying model/algorithm, easily identify that bird and learn more about different species along with their sounds. Through this app, a child would be able to:

  • Explore different species of birds and their corresponding sounds
  • Record new bird sound and identify the species through our database
  • Increase their awareness of local fauna
  • Increase their awareness about birds not found in their locality We are constantly preparing labelled data to incorporate as many species as we can, to give a near exhaustive list of species to learn from. The database will be regularly updated even after the app is ready, creating a one-stop app for comprehensive knowledge of the local fauna.

Project Plan:

Step Know-how Status
Data collection - collect sounds of different species of birds (labeled data) Android Development In-Progress
Develop and test Deep Learning Models to classify sounds of Birds into their corresponding species Machine Learning and Deep Learning Algorithms In-Progress
Conceptualize and integrate the designed ML model with an Android app. To make API call to get detailed info about the species found (Deployment) Android Development Not Started