AI Detects Smuggled Marine Wildlife with 92% Accuracy

An artificial intelligence (AI) algorithm can identify smuggled marine wildlife with 92% accuracy. The system targets shark fins, seahorses, and sea cucumbers. These items are frequently trafficked across borders. This technology could enhance efforts to combat illegal wildlife trade.

Researchers at Macquarie University developed the algorithm. They trained it using X-ray computed tomography (CT) scans. These scanners are common at airports for security checks. The AI learns to recognize specific patterns of trafficked marine species within these scans. This method aims to automate the detection process.

The team created 298 scans from 20 sea cucumber, 30 seahorse, and 18 shark fin samples. Many samples came from previous wildlife seizures. They scanned items in various positions and contexts. Some scans mimicked smuggling tactics, like hiding items in foil or toys. The algorithm was then tested on new images it had not seen before.

Overall, the algorithm achieved a 92% success rate. It detected shark fins with 95% accuracy and seahorses with 96% accuracy. Sea cucumbers were identified with 86% accuracy. The false positive rate was 13%, with 2% for shark fins, 1% for sea cucumbers, and 9% for seahorses. This high accuracy suggests the AI can significantly aid detection efforts.

This AI system complements existing detection methods. It is not a complete replacement for human inspectors or sniffer dogs. Not all airports have access to 3D CT scanners. The technology offers a powerful tool to intercept illegal shipments. It can help secure convictions against marine wildlife traffickers. This effort supports the conservation of marine ecosystems.