Hi,

I am jason Holmberg from WildMe.org. I am one of the developers of Wildbook (wildbook.org), an open source data management platform for wildlife research. I'm using ML as part of the IBEIS.org project to boost and metascore multiple computer vision algorithms for individual humpback and sperm whales. David, I would love to speak offline if you have the time: [email protected].

Cheers,

Jason

Hi Steph,

Thanks for your message. I would be delighted to answer your question about Wildsense.

You are correct that our aim is to do a lot more than raise awareness and engange citizens with our Wildsense Tigers game.

We are using images from a range of image sharing websites that have been uploaded by people online and publicly shared. Each photograph contains valuable information that often includes the time and location that the photograph is taken. If we can understand when a photo was taken, where it was taken, and the individual animal in the photograph, then we can start to build up a picture of the movement of individuals. We are benefiting from people power from the users that are both indirectly uploading photographs online and also directly interacting with our Wildsense game. We are combining this with computing power with our data analysis, computer vision and machine learning software.

We are doing this experiment on a large scale and we believe that "the wisdom of the crowd" will enable us to gather and analyse enough information so that we can build an accurate picture. We started with wild tigers and our plan is to create a platform that can be used for other species as well. We are working on other animals already behind the scenes.

I hope that answers your question.

Best wishes,

Aaron Mason

Hi Barbara,

You are right in that cyberpoaching is becoming an increasing risk. I have seen various cases where information posted on social media has provided too much information and then ended up in the wrong hands (e.g. http://www.army.mil/article/75165/Geotagging_poses_security_risks/).

We need to be careful about the accuracy of information that is disclosed. Providing an approximate time and location seems to be the most popular adopted approach. 

Hacking is another issue. For example, I have seen research demonstrating the potential to gain unauthrised access to GPS collar data.

Our main focus has been the Bengal Tiger but we want to see how much we can learn about other tigers as well. Our end goal is to support a range of animal species.

P.S. I am pleased that you are aware of http://www.tigernation.org - I co-founded it :)

"TensorFlow, you see, deals in a form of AI called deep learning. With deep learning, you teach systems to perform tasks such as recognizing images, identifying spoken words, and even understanding natural language by feeding data into vast neural networks. "

Would this be applicable to an acoustic monitoring network? For example. my research has shown tigers have unique, identifiable vocalizations down to the individual and sex. If this software is applied to my recording network for tigers, would it be able to automatically recognize and categorize these individuals?

For example: when it hears Tiger 108, it would know and then input that it heard Tiger 108 at a particular time and date.

The catch will be (and for any neural network or AI type learning I would expect the same) the training phase. If you are able to tell the sounds apart or identify a specific sound as belonging to a certain individual, the AI should afterwards be able to automatically identify the critical factors needed to distinguish the voices of the individuals. But it will need enough input from each individual as well as the different vocalizations used by tigers. AFAIKT it will be able to do this automatically afterwards, but I am not sure if (a) you will get enough identifiable vocalisations and (b) with a wide enough range of typical tiger vocalisations for it to be really reliable. Training on zoo animals might work? I am also interested in this, but for jackals instead of tigers.

I'd like to suggest our open source package Wildbook (http://www.wildbook.org) as a base data management platfor for this. I agree with the above that there are a number of challenges around the vocalizations themselves, but having the identity information in a good database and data model is a great foundation. That's what we're doing for our computer vision/deep learning project at www.IBEIS.org.

Our non-profit WildMe.org is running both. Feel free to contact us with questions. We have played with time series matching (often used for speech recognition)...but actually for whale flukes. Would be happy to discuss potential for audio ID. 

Thanks for sharing this, Sarah.

Hi Dave,

I believe with any DJI drones you can plot out the course of a drone and let it fly along the route with you controling the speed and height of the drone. The limitation is that it has to eb within a mile of the pilot.

Regards,

John

Thank you for sharing!

Oh wow. This is perfect for what I do... tracking tigers by their voices alone!