Pakheta

• 

  Landing

• 

  Bird Habitat Map

• 

  Spectrogram Canvas

• 

About the project
Pakheta is a bird conservation project built to use advanced AI techniques for monitoring and protecting endangered bird species in Nepal. By analyzing bird calls, it helps preserve biodiversity and habitat.

Inspiration

Our inspiration for Pakheta came from the alarming rate at which bird species are becoming endangered in Nepal. With nearly 20% of the bird population at risk, including species vital to the ecosystem, we saw a pressing need for more efficient, real-time conservation tools. The idea of using AI to analyze bird calls and track their populations was born from the desire to make conservation more accessible and actionable.

What it does

Pakheta uses AI, specifically Convolutional Neural Networks (CNN), to identify bird species based on audio recordings of their calls. By analyzing spectrograms of these calls, the system can monitor bird populations in real time, even in remote areas that are hard to access. This technology enables faster and more accurate tracking, providing critical data for habitat protection and conservation strategies.

How we built it

The project was built using a combination of machine learning techniques and cloud technologies. We trained our AI models using a large dataset of bird call recordings, employing MS (Mel Spectrogram) to process the audio and CNNs for classification. The platform provides real-time monitoring through an easy-to-use interface for researchers, conservationists, and the general public.

Challenges we ran into

One of the biggest challenges we faced was gathering enough high-quality bird call data to train our models effectively. Additionally, building an AI system capable of distinguishing between similar bird calls in noisy environments required fine-tuning our models. Ensuring that the system was scalable and could operate in remote areas with limited resources was another technical hurdle.

Accomplishments that we're proud of

We’re proud of successfully developing an AI system that can identify bird species from audio recordings with high accuracy. The real-time monitoring capability allows conservationists to make more informed decisions about habitat protection and species conservation. Additionally, we are excited about how this project can be scaled to benefit conservation efforts globally.

What we learned

Through this project, we learned how AI can play a transformative role in conservation. We also gained a deeper understanding of the challenges involved in working with environmental data and how to overcome those hurdles. We improved our skills in machine learning, audio processing, and deploying AI models in real-world scenarios.

What’s next for Pakheta

We plan to expand the project by incorporating more bird species and improving the AI model's accuracy. Our next steps involve developing a mobile app for easier access and expanding our user base to involve local communities in bird conservation efforts. We also aim to collaborate with local organizations and government agencies to enhance the impact of Pakheta and drive meaningful change for Nepal’s bird populations.

Built With

• django
• hugging-face
• react
• tensor-flow