Case Study: Rainforest Connection improves audio spectrogram visualization with Clockwise Software

Rainforest Connection helps stop illegal logging with Clockwise Software

Rainforest Connection, an environmental startup, needed to visualize vast amounts of audio recordings from forests and oceans to identify animal sounds and detect illegal activity like deforestation. Their challenge was to build a web-based tool that would allow their team to manually select and annotate specific segments on audio spectrograms to train a machine learning model. They hired Clockwise Software for its D3.js development expertise to create this functionality and integrate it into their existing infrastructure.

Clockwise Software implemented a solution using D3.js and Canvas to create interactive, zoomable spectrograms synced with audio playback. The key features included a main spectrogram, a macro overview spectrogram, and tools for creating and managing both manual annotations and AI-generated detections. As a result, the system successfully trained AI to identify chainsaw sounds, leading to the apprehension of illegal loggers in Malaysia. Furthermore, the tool is now used by scientific institutes to study whale and monkey behavior. This business success helped Rainforest Connection secure investment from Google.

Rainforest Connection