A group from Auckland University has been training AI models using our thermal video library.
A group from Auckland University has been training AI models using our thermal video library.
In the area of monitoring, we suspect that Cacophony’s thermal cameras have the potential to deliver three keys benefits:
Earlier in the year we were asked if we could expand the machine vision used with our thermal camera to automatically detect wallabies with the goal of monitoring and controlling the wallaby population.
If you're a regular reader of our blog, it might be easy to get the impression that we are randomly trying lots of things. The reality of life on the project is somewhat different - there is a targeted structure to what we are doing. This post gives a summary of that strategy. To bring back the Cacophony of native fauna in NZ there are a number of separate parts of the puzzle that we are trying to solve. What follows is a summary of each sub-goal as we see it along with an indication of where we are up to with our progress.
Our previous blogs have highlighted how most predators in well trapped areas just walk past existing traps. This blog shows our first attempt at a device designed to trap hard to trap predators or re-invading predators. At the moment the competition for this sort of application is pretty much manual hunting or very intense trapping and baiting (which often never gets to zero).
In recent blog posts we have gone into detail on different trapping strategies. We chose to do this because we are convinced that many approaches that may seem intuitively to be great strategies won’t actually make much of a difference to the elimination of predators. Our camera experiments have shown consistently and across a number of different environments that, for an area that has been trapped for a while, there is a persistent population that avoids existing traps. Today we introduce a collection of approaches and ideas that we believe can actually improve the predator interaction rate and give us a real chance of achieving our predator-free goals.
Today we tackle the question of the kill rate of existing traps. The arsenal of traps available to trappers includes some well-designed, field-tested, and hardy workhorses. And yet we know that even the most skillful deployment of these in the field only delivers a level of suppression, not the total elimination we strive for. Today we discuss why that might be.
The previous blog posts showed how a simple model can help understand typical predator elimination methods. It makes intuitive sense that long life lure or automatic dispensing lures will help with trapping. In this blog we discuss the impact of automatic lures and long life lures for their potential to achieve total predator elimination.
Today we return to our core subject of eliminating predators. If you're keen to follow our work on the fever screening device (which is now on the market), we have now launched a new website with its own blog: http://www.tekahuora.com. All future updates on the device will appear there.So, back to dealing with predators. In some previous blog posts, we showed how a simple model can help understand typical trapping and poisoning methods.
As you may have noted from our recent blog entries life is getting busy for us here at The Cacophony Project. And you know we love building solutions right? So, we've been building really quite a lot of software recently and we need some help making sure it makes it out of the door in the highest quality possible.
Regular readers of this blog will have noted our recent pivot in focus. Our team are busy working on our thermal screening device and the devices are already out at Beta testing sites helping employers keep their staff safe. We're pleased to share how the media have been taking notice of our efforts. And a look ahead to how we see our solution fitting in with some of the innovations that are being designed to help keep New Zealand safe from further outbreaks.
The Cacophony Project is all about the eradication of invasive predators from New Zealand. However, there's a wee thing called COVID-19 that has brought about a few problems in human society (you may have seen the odd thing about it in the news recently). At Cacophony, we love being useful so when the call came requesting that we turn our attention to how our cameras might be used, we listened and responded with our favourite type of action - we started building things. Today, we talk to our founder about how that happened and what we're doing to try to help New Zealand get through this and return to something like normal.
The Eastern Bays Songbird project has been going for 2 years and the members were keen to see how they were getting on with their goal of predator eradication. They have been using a 2040 Thermal Camera developed by The Cacophony Project to see what is going on around their traps. Their overall impression before deploying the camera was that they were getting the predator numbers down. They hoped the Thermal Camera would be a great way to get a full picture of the predators in the area. This camera was developed because standard trail cameras are designed for pig and deer and miss lots of the small moving mammals we are interested in here. In this entry, we share their findings.
Our previous blog post showed a simple model to allow you to work out trap interaction rates. The goal of this post is to show that this same simple model can also be used to explain the effectiveness of aerial poison drops. We are not making any comments about the value of poison drops as a tool but just showing how a simple model works for different elimination methods. We like the fact that a fairly simple tool can help explain the relative merits of different elimination methods. The real value starts to appear when we tweak some of the other parameters in the model - we think it gives us a clear idea about useful ways to improve predator elimination.
Over the last three years we have had various versions of our thermal cameras in front of all the main types of traps. What we have found is that a large number of predators seem to just walk on past irrespective of what type of lures and set ups we use. We have detailed videos of rats and possums running all around traps but not always interacting with them. Typical reaction to this is that we must be doing something wrong but after watching tens of thousands of videos we suspect this is more of an issue than just us being hopeless trappers.
Today we introduce a model we have developed to allow you to calculate the interaction rate of your traps. Given inputs such as interaction rate, elimination rate, number of devices, and predator population the model creates a graph showing the likely impact on the population. We think the results are not only intriguing but telling. They suggest it might be time for all of us to adjust our view of the importance of some of the factors involved.
Self-resetting (automatic) traps have been on the market for a while now. Speaking to trappers far and wide (as we have a habit of doing) we hear mixed reviews. Most people seem to get some success when first deploying an auto-trap but the results tend to dip pretty swiftly. At Cacophony, we do value such anecdotal evidence (it really helps us understand the problems of using devices in the field) but we value hard, physical evidence even more.
The first version of the Cacophony Index is now live. The Cacophony Project was first started to attempt to answer a seemingly simple question: "is a bird population getting better or worse over time?". We now have a tool available which can start to answer this question.
We recently came across a new and very cool piece of research which investigated the effectiveness of conventional trial cameras for detecting various North American mammals. To conserve power, trail cameras stay asleep most of the time and use a basic motion sensor to wake the camera up to take photos. To test how many animals were missed by the motion sensor they configured the cameras to take a photo every minute and also whenever the motion sensor detected triggered.
Author: David Blake
David Blake is a semi-retired property investor who now likes to spend his time trapping pests and planting native trees. From time to time he helps out by volunteering at The Cacophony Project doing some filming and occasionally contributing to parts of the software.
Conventional trail cameras are cheap and offer high image resolutions. They are designed for detecting larger animals such as pigs and deer. Thermal cameras, like those used by the Cacophony Project, are much more expensive and typically have lower resolution. Conventional cameras need an IR light source to "see" at night while thermal cameras do not.
Some of the predator monitoring and trapping solutions we are developing require more power than traditional tools. This article explores why we think the effectiveness of tools is more important than power use for most of the applications we are interested in.
We're really excited about 2040, a social venture that is commercialising the technologies developed by the Cacophony Project. They've recently published a blog article about some ongoing research by Tim Hunt from Wintec who is using audio recordings gathered by the Bird Monitor product to automatically detect and identify morepork calls.
We have created a new tool that allows us to test many different methods to lure and capture predators of any type, using sound and thermal vision. We are calling this tool the Cacophony Predator Lab.
In this blog article, we will discuss how sound lure experiments can be set up to test the effectiveness of various sounds for attracting possums. The sound lure software is configured from a web interface that allows you to upload any set of sounds and play them at any volume or time sequence.
Author: Ben McEwen
Hi, I’m Ben. I am a student in my final year of Mechatronics Engineering at the University of Canterbury. This semester I have been working with the Cacophony Project to explore improved ways of tracking animals in video footage for better predator recognition and for the elimination of invasive predator species.
Recent testing of the Cacophony Project's thermal camera has reinforced how good this type of technology is for predator monitoring and control. This blog post highlights the core reasons that the thermal camera developed by The Cacophony Project is a great technology for this application.
The key advantages are:
As highlighted in earlier blog posts, the Cacophony Thermal camera is much more sensitive for the detection of predators than the next best tool. It is able to see 3 to 50 times more activity, allows us to make observations that have never been possible before.
Last week, Menno Finlay-Smits from the Cacophony Project participated on a Techweek 2019 panel which was centred on the idea of "Tech for Good". The panelists each represented an organisations that each attempt to "do good" in the world in its own way. We discussed various ways that organisations can do this, as well as emerging entity structures that focus on more than just profit.