Singapore’s marine biodiversity has found a new digital ally as a group of local youths and volunteers launch SGFish, a free mobile application designed to help recreational anglers identify and log over 900 fish species found in local waters. Developed by members of the Ichthyological Society of Singapore, the app debuted on the Apple App Store on September 17, combining advanced computer vision with citizen science to protect vulnerable marine life.
Despite busy shipping lanes, Singapore sits near one of the globe’s most biodiverse marine regions. That geographical quirk yields a surprising variety of marine life, though their ecological significance often escapes casual fishermen. A striking example occurred in August at Bedok Jetty, when an angler mistook a globally endangered zonetail butterfly ray for an ordinary stingray. Brian Sng, a 21-year-old software engineering student at Singapore Management University, reviewed a photograph of the catch and quickly alerted the fisherman to the rare find.
Building SGFish to Combat Mistaken Identities and Protect Marine Life
That close call catalyzed a tech-driven response. Sng joined forces with fellow enthusiasts to build a tool addressing the frequent issue of mistaken identities among local fishers, who sometimes confuse dangerous species with harmless ones or overlook sustainable practices. The mobile application features a built-in pop-up alert system advising users on proper protocols when they scan protected, rare, or harmful species, explicitly prompting the immediate release of endangered catches.
To ensure high identification accuracy, Sng served as lead developer, collating and vetting more than 90,000 photographs of fish species native to Singapore and neighboring waters. Sourcing images from international databases, society members, and his own archives, he spent two months training the app’s computer vision model. The resulting algorithm achieved an 82.7 percent accuracy rate, outperforming global flora and fauna databases like iNaturalist, which typically hover around 80 percent accuracy for fish identification worldwide.
From Telegram Bot Roots to a Comprehensive Mobile Encyclopaedia
The initiative actually began as a simpler Telegram bot named Angler’s Pokedex, a nod to the electronic encyclopaedia featured in the Pokémon franchise. Samuel Pua, a 36-year-old cybersecurity professional and volunteer developer, created the original bot to compile community catches. However, limitations regarding automated time and location extraction from Telegram images prompted the team to transition toward a dedicated smartphone application.
The newly launched SGFish app refines that data collection process by automatically logging geographic coordinates alongside user-submitted photographs. To maintain rigorous scientific integrity, society volunteers and co-founders manually verify every submitted record before it enters the official database. Currently, the repository contains approximately 2,300 observations spanning over 400 species, building on data first gathered by Pua’s Telegram bot.
Overcoming Murky Waters With Lightweight Offline Tech and Tide Data
Field usability drove several key design choices during development. Guan Chen Di, a 22-year-old computer science student at Nanyang Technological University and fellow developer, noted the complex engineering challenge of condensing comprehensive marine data onto small mobile screens. The team successfully streamlined the user interface to display the image, species data, and catch method within a single cohesive view.
To mitigate the risk of classification errors, the software displays the top five predicted species matches, allowing the user to select the correct identification manually. For anglers operating in areas with poor cellular reception, the application includes a lightweight offline model for rapid field assessments, supplemented by a more comprehensive online version. The app also integrates tide data and a digital fish encyclopaedia to engage recreational fishers.
Fish biologist Jeffrey Kwik noted that citizen science platforms like SGFish provide invaluable supplemental data for traditional marine surveys, which frequently struggle in Singapore’s notoriously murky waters where visual diving and snorkelling counts remain challenging. Future updates for the platform will introduce a Google Play Store release and reintroduce gamification elements from the original Telegram bot, including high-score boards honoring users who document the highest diversity of unique species.
