An interactive web tool lets users recognize a London Underground line solely from its auditory signature. Visitors type or play a clip, and the system matches it against a curated database of train announcements, station chimes, and rolling‑stock timbres.

The project showcases practical AI‑driven classification: audio samples are converted into spectrograms, then fed to a lightweight convolutional neural network trained on thousands of labeled clips. The result is a 92‑percent accuracy rate reported by the creator.

Beyond novelty, the tool offers accessibility gains for visually impaired commuters who rely on sound cues to navigate the Tube. Developers can adapt the open‑source code for other transit systems or for educational sound‑recognition modules.

Ultimately, the site demonstrates how small‑scale machine learning can transform everyday navigation into a richer, more inclusive experience.