/portfolio

AI Bird Sound Recognition

Scope:

Develop automated animal vocalization recognition technology

Industry:

BioTech

Customer:

NDA

Timeline:

The project span a 5-year period, beginning in 2021 and scheduled for completion in 2025

Team

Machine Learning / Deep Learning Engineers

Mobile Developers

QA Engineer

Key technologies

TensorFlow

Keras

Librosa

Docker

REST API

Mobile platforms: iOS, Android

Problem

Need for continuous monitoring of endangered species and biodiversity in forest ecosystems

GOAL

Create an AI-powered system capable of recognizing and categorizing bird species by their vocalizations

SOLUTION

An automated system to capture and analyze bird vocalizations
Integration with mobile (iOS and Android) and web platforms for real-time monitoring
Continuous data collection and ecosystem monitoring, ensuring effective conservation efforts

Impact

2300+ bird vocalization recordings

Annotated with 41,000+ labels covering 116 species

85% boost in user engagement

Making conservation efforts more interactive and community-driven

50% increase in shared content with community