Industry:
BioTech

Customer:
NDA

Task:
Develop an automated animal vocalization recognition technology for continuous autonomous monitoring of rare, endangered, and indicator species, as well as biodiversity in forest ecosystems.

Team:
3 engineers specializing in Machine Learning (ML) and Deep Learning (DL),
1 BackEnd developer,
2 mobile developers (iOS and Android),
1 web developer,
1 QA engineer,
1 designer,
1 project manager.
Technologies: Tensorflow, Librosa, Pandas, NumPy, Keras, Django, MySQL, REST API, Docker, Nginx/Apache, SSL/TLS, Git.

Timeline:
The project is planned for 5 years, starting on 2021, and ending on 2025.
Currently, data collection, voice signal annotation, and the development of automated recognition algorithms are complete. A software prototype is in place for Web, Android, and iOS versions.

Solution:
We developed a dataset of over 2,300 bird vocalization recordings, which were carefully annotated with more than 41,000 annotations for 116 bird species. The technology enables automated voice signal recognition and annotation, integrated into an information-analytical center (IAC). The IAC is accessible through Telegram bots, mobile apps, and a web version. Access varies for general users and researchers.

Benefits:

  • Enhanced Learning Experience: Users can actively engage with the AI, improving their knowledge of bird species while enjoying interactive birdwatching.
    • 85% increase in user engagement with interactive features.
  • Increased Accessibility: The platform supports multiple languages, making it usable for a wider audience and promoting inclusivity in biodiversity education.
    • 70% growth in user base due to multilingual support.
  • Convenient Access: Available as mobile apps and web platforms, users can easily interact with the system anytime and anywhere.
    • 90% user satisfaction rate for accessibility and ease of use.
  • Encouragement of Conservation Efforts: By raising awareness of local biodiversity, the platform encourages users to participate in conservation initiatives and protect natural habitats.
    • 60% of users reported increased participation in conservation activities.
  • Community Engagement: Users can share their experiences and knowledge, fostering a community of nature enthusiasts and promoting collaborative learning.
    • 50% increase in community interactions and shared content.
Contact us