I see you, Fish Face

I don’t know how accurate scenes were in the latest spy thriller movie I watched, but I think the part where they used facial recognition software to identify a terrorist was pretty close to reality.

A closed circuit camera caught a glimpse of the bad guy’s face as he snuck into a supposedly secure facility. The computer wizards at the CIA dumped the photo into their computer and it almost instantly scanned millions of possibilities and came up with the identity of the perpetrator.

Wow! A grainy glimpse of a face was as good as a finger print.

Hopefully, this type of artificial intelligence will soon be available to fight a different but pervasive type of crime - illegal commercial fishing. Rather than picking out faces, the software will track the movement of fishing boats to root out illegal behavior. As important, using a twist on facial recognition, the computer will be able to recognize when a boat's haul includes endangered or protected fish.

It’s no secret I’m no fan of commercial fishing. I would have little disgust with it if all commercial fishermen operated above the law. I’d accept it if by-catch, where non-target species of fish are caught, was rare or eliminated. I’d be okay with it if species commonly sought by recreational fishermen were off limits or at least the allocation between sportsmen and commercial netters was fair. I’d like it if commercial fishing methods injurious to the natural environment were eliminated. I believe many species of fish could provide a sustainable harvest but in almost every case, commercial fishermen operate well past what’s sustainable.

Too bad few, if any, of my contentions are met, inland, on the Great Lakes or in salt water. Most fisheries managers will secretly or openly agree with me and efforts to correct the abuses of commercially harvesting wild fish are on going.

An effort to use artificial intelligence to fight illegal fishing is coming from Virginia-based The Nature Conservancy (TNC). TNC recently launched a contest on Kaggle - a crowdsourcing site based in San Francisco using competitions to advance data science.

TNC hopes the winning team will produce computerized cameras loaded with software capable of identifying specific species of fish much the same as facial recognition programs identify spies in the movies and then documents the catch being hauled aboard.

Currently, some commercial vessels are already required to have cameras videoing what’s being caught. However not every video actually gets watched since doing so is very time consuming. Even in the videos being watched, it’s impossible for the human viewer to positively ID many of the fish being caught.

It’s hoped the software will do a better job of identifying what’s being caught and can then put an electronic marker at each point in the video when a protected fish is hauled out. Inspectors will then be able to go directly to those moments and check a fishing crew's subsequent actions to determine whether they handled the catch legally.

TNC expects this approach could cut review time by up to 40% and increase the actual monitoring on commercial boats. Despite rules that call for government-approved auditors to be stationed on commercial fishing boats in some areas; in practice, catch auditors are found on only 2% of the fishing boats in salt water. There’s no requirement for commercial boats on the Great Lakes, tribal or otherwise, to have spotters aboard or even stationed at the docks to see what’s off loaded.

It’s estimated world wide, illegal, unreported and unregulated fishing represents up to 20% of all seafood. There’s no estimate about IUU fishing on the Great Lakes, but it certainly exists. "If using facial recognition software to track fish were easy, we'd already be using it," says Matthew Merrifield, TNC's chief technology officer. “Whereas images from security cameras installed inside banks or other buildings are consistent and predictable, the data from cameras on boats is inconsistent because the ships are always moving and the light keeps changing."

The winning team of this contest will earn a prize of $150,000. Then, as part of its campaign to reduce bycatch and illegal fishing, TNC will work with the governmental agencies to install the equipment and software for free on selected fishing boats.