Camera traps have been a key part of the conservation toolkit for decades. Remotely triggered video or still cameras allow researchers and managers to monitor cryptic species, survey populations, and support enforcement responses by documenting illegal activities. Increasingly, machine learning is being implemented to automate the processing of data generated by camera traps.
A recent study published showed that, despite being well-established and widely used tools in conservation, progress in the development of camera traps has plateaued since the emergence of the modern model in the mid-2000s, leaving users struggling with many of the same issues they faced a decade ago. That manufacturer ratings have not improved over time, despite technological advancements, demonstrates the need for a new generation of innovative conservation camera traps. Join this group and explore existing efforts, established needs, and what next-generation camera traps might look like - including the integration of AI for data processing through initiatives like Wildlife Insights and Wild Me.
Group Highlights:
Our past Tech Tutors seasons featured multiple episodes for experienced and new camera trappers. How Do I Repair My Camera Traps? featured WILDLABS members Laure Joanny, Alistair Stewart, and Rob Appleby and featured many troubleshooting and DIY resources for common issues.
For camera trap users looking to incorporate machine learning into the data analysis process, Sara Beery's How do I get started using machine learning for my camera traps? is an incredible resource discussing the user-friendly tool MegaDetector.
And for those who are new to camera trapping, Marcella Kelly's How do I choose the right camera trap(s) based on interests, goals, and species? will help you make important decisions based on factors like species, environment, power, durability, and more.
Finally, for an in-depth conversation on camera trap hardware and software, check out the Camera Traps Virtual Meetup featuring Sara Beery, Roland Kays, and Sam Seccombe.
And while you're here, be sure to stop by the camera trap community's collaborative troubleshooting data bank, where we're compiling common problems with the goal of creating a consistent place to exchange tips and tricks!
Header photo: ACEAA-Conservacion Amazonica
WCS is seeking a Conservation Technology Specialist to join their work in the Okapi Wildlife Reserve.
11 August 2023
Please join us in celebrating this year’s top #Tech4Wildlife Photo Challenge Honorees as chosen by our panel of leading conservation organization judges, and enjoy the story contained within these entries about how our...
4 August 2023
Join us as we count down the WILDLABS community's honorees in the first-ever #Tech4Wildlife Community Choice Awards!
3 August 2023
Exciting opportunity for an experienced biodiversity monitoring expert in ZSL's conservation department
18 July 2023
Article
In 2019, the U.S. Navy initiated a time-lapse camera study to investigate seal presence at select haul-out locations in the lower Chesapeake Bay and coastal waters of Virginia, which are important areas to Navy training...
13 July 2023
Apply for funding (£500,000-£750,000) to develop software systems, which will help to improve biodiversity monitoring by automating the analysis of images and videos
12 July 2023
Applications are now open till 23 July for the 2023 Canon Oceania Grants with the category of the Environmental Grant valued at AU$5,000. The finalist will be selected based on the environmental and social merits of...
7 July 2023
Article
At Appsilon, we are always working to enable our users to get the most out of our solutions. With this in mind, we are happy to introduce two new add-ons to Mbaza AI.
4 July 2023
In this Conservation Tech Showcase case study from Osa Conservation, you’ll learn about how technology is aiding their long-term efforts to prevent wildlife crime, protect critical species, and build a climate-adaptive...
22 June 2023
In this interview between Rob Appleby and Ellie Warren, we discuss the importance of DIY, recycling, and sharing tools in order to make our community more sustainable.
14 June 2023
Careers
Permanent and Full Time role at the UK Centre for Ecology & Hydrology
8 June 2023
Acorn removal study of Nendo Dango, Ecological Restoration Research group at the University of Granada
5 June 2023
April 2024
event
October 2024
December 2023
event
November 2023
event
Using "motion extraction" for animal identification
16 January 2024 3:46pm
17 January 2024 5:16am
You are most welcome.
I can only highly recommend joining the online events! They are really inspiring.
- and it is great to "meet" and chat with fellow WildLabbers during the "after hours".
17 January 2024 4:22pm
Very nice video in the link you posted btw:
Here is another less artistic one:
Wildlife Conservation for "Dummies"
9 January 2024 10:02pm
10 January 2024 11:24pm
Maybe this is obvious, but maybe it's so obvious that you could easily forget to include this in your list of recommendations: encourage them to hang out here on WILDLABS! I say that in all seriousness: if you get some great responses here and compile them into a list, it would be easy to forget the fact that you came to WILDLABS to get those responses.
I get questions like this frequently, and my recommended entry points are always (1) attend the WILDLABS Variety Hour series, (2) lurk on WILDLABS.net, and (3) if they express a specific interest in AI, lurk on the AI for Conservation Slack.
I usually also recommend that folks visit the Work on Climate Slack and - if they live in a major city - to attend one of the in-person Work on Climate events. You'll see relatively little conservation talk there, but conservation tech is just a small subset of sustainability tech, and for a new person in the field, if they're interested in environmental sustainability, even if they're a bit more interested in conservation than in other aspects of sustainability, the sheer number of opportunities in non-conservation-related climate tech may help them get their hands dirty more quickly than in conservation specifically, especially if they're looking to make a full-time career transition. But of course, I'd rather have everyone working on conservation!
13 January 2024 3:14am
Some good overview papers I'd recommend include:
- Besson, M., Alison, J., Bjerge, K., Gorochowski, T. E., Høye, T. T., Jucker, T., ... & Clements, C. F. (2022). Towards the fully automated monitoring of ecological communities. Ecology Letters, 25(12), 2753-2775.
- Speaker, T., O'Donnell, S., Wittemyer, G., Bruyere, B., Loucks, C., Dancer, A., ... & Solomon, J. (2022). A global community‐sourced assessment of the state of conservation technology. Conservation Biology, 36(3), e13871.
- WILDLABS-led research! Led by @TaliaSpeaker and @StephODonnell
- Lahoz-Monfort, J. J., & Magrath, M. J. (2021). A comprehensive overview of technologies for species and habitat monitoring and conservation. BioScience, 71(10), 1038-1062.
- Tuia, D., Kellenberger, B., Beery, S., Costelloe, B. R., Zuffi, S., Risse, B., ... & Berger-Wolf, T. (2022). Perspectives in machine learning for wildlife conservation. Nature communications, 13(1), 792.
- Stowell, D. (2022). Computational bioacoustics with deep learning: a review and roadmap. PeerJ, 10, e13152.
- I'm biased towards bioacoustics because that's what I focus on, but this regardless this is an excellent horizon scan of AI/ML for conservation and acoustics!
- Borowiec, M. L., Dikow, R. B., Frandsen, P. B., McKeeken, A., Valentini, G., & White, A. E. (2022). Deep learning as a tool for ecology and evolution. Methods in Ecology and Evolution, 13(8), 1640-1660.
I'd also encourage you to follow the #tech4wildlife hashtags on social media!
15 January 2024 4:27pm
I'm also here for this. This is my first comment... I've been lurking for a while.
I have 20 years of professional knowledge in design, with the bulk of that being software design. I also have a keen interest in wildlife. I've never really combined the two; and I'm starting to feel like that is a waste. I have a lot to contribute. The loss of biodiversity is terrifying me. So I’m making a plan that in 2024 I’m going to combine both.
However, if I’m honest with you – I struggle with where to start. There are such vast amounts of information out there I find myself jumping all over the place. A lot of it is highly scientific, which is great – but I do not have a science background.
As suggested by the post title.. a “Wildlife Conservation for Dummies” would be exactly what I am looking for. Because in this case I’m happy to admit I am a complete dummy.
HWC Tech Challenge update 2020
Testing Raspberry Pi cameras: Results
5 May 2023 5:11pm
5 September 2023 8:16am
And finally for now, the object detectors are wrapped by a python websocket network wrapper to make it easy for the system to use different types of object detectors. Usually, it's about 1/2 a day for me to write a new python wrapper for a new object detector type. You just need to wrap in the network connection and make it conform to the yolo way of expressing the hits, i.e. the json format that yolo outputs with bounding boxes, class names and confidence level.
What's more, you can even use multiple object detector models in different parts of a single captured image and you can cascade the logic to require multiple object detectors to match for example, or a choice from different object detectors.
It's the perfect anti-poaching system (If I say so myself :) )
10 January 2024 11:47pm
Hey @kimhendrikse , thanks for all these details. I just caught up. I like your approach of supporting multiple object detectors and using the python websockets wrapper! Is your code available somewhere?
11 January 2024 6:02am
Yep, here:
Currently it only installs on older Jetsons as in the coming weeks I’ll finish the install code for current jetsons.
Technically speaking, if you were an IT specialist you could even make it work in wsl2 Ubuntu on windows, but I haven’t published instructions for that. If you were even more of a specialist you wouldn’t need wsl2 either. One day I’ll publish instructions for that once I’ve done it. Though it would be slow unless the windows machine had an NVidia GPU and you PyTorch work with it.
SDZWA Conservation Tech Summer Fellowship
9 January 2024 7:04pm
ReWilding Europe a Data Annotation Pipeline for training models [code and model weights]
2 January 2024 6:05am
Open-Source design guide for a low-cost, long-running aquatic stereo camera
13 December 2023 5:47pm
15 December 2023 4:53pm
This is awesome - thanks for sharing Stephanie!! We actually were looking around for a low-cost video camera to augment an MPA monitoring project locally and this looks like a really great option!
Cheers, Liz
28 December 2023 8:54pm
Thank you for sharing! Super interesting, as we don't see many underwater stereo cameras! We also use Blue Robotics components in our projects and have found them reliable and easy to work with.
Apply to Beta test Instant Detect 2.0
11 May 2023 10:55am
21 December 2023 3:48am
Will you accept personal/hobbyist focused on conservation on their small plots of land (10-100 acres)?
I would, and know others, who would happily pay more than the official conservationists rate for the service, which could help to further subsidize the project. (Referring to your statement here: https://wildlabs.net/discussion/instant-detect-20-and-related-cost)
Mesh camera trap network?
9 April 2017 1:29pm
8 November 2021 1:05pm
Hi Sam,
Impressive! Any chance the LoRa code is open source? I should like to take a gander.
Thanks
21 December 2023 3:41am
How's the project going? Plans to open-source and share more about it?
The website mentions it's in alpha stage, going into beta.
21 December 2023 3:42am
Oh I see recent threads on the topic. Sharing here to link people over.
Subsea DIY Burnwire for Deep-sea BRUVS
6 December 2023 3:49am
11 December 2023 10:21am
Yeah from memory we found it difficult to get the relatively high voltage (~50VDC) and current (can't remember) in a small package, but we had almost no experience back then and gave up fairly quickly. We also found it difficult to get much help from the company if I remember correctly...
so is the problem with the nichrome waterproofing everything? I picture something like coating the nichrome in high temp grease (especially where it's in contact with the nylon line and the line itself) and encapsulating the entire thing in a semi-flexible silicone (so the line can slip out after detechment) with something buoyant to help pull it towards the surface maybe? Speaking of, how are tags being recovered (i.e. do they need to pop to the surface)?
15 December 2023 4:17pm
Hi Titus,
We've used this design/procedure for many years with our Deep Sea Camera systems, with good reliability. Not OTS but not hard to make and most of the materials come out to be inexpensive per unit. The most expensive item is the M101 connector ($25ea), but if you get them with extra length on the cable, you can essentially cut it off at the point where it joins the burn-loop and reuse that connector until it gets too short. You'd also need an F101 connector integrated with your BRUV, this connecting with the burnwire and forming the the positive side of the circuit, and a ground - our ground connection goes to a large bolt on the frame near the burnwire loop - but that connector generally shouldn't need replacement unless it gets damaged.
These burnwires generally break in 3-7min, burning at about 1Amp, ~14.5V. A thinner version of the coated wire could go faster or with less power required.
We do also employ galvanic releases as backups. I really like redundancy on recovery mechanisms! The ones we use are made by International Fishing Devices, Inc. Various distributors sell certain models of their products (i.e. different time durations) but if you contact them directly, they can also make custom duration ones for you.
19 December 2023 4:47pm
Hi Titus,
I've used latching solenoids as a release in a fresh water application. The product linked to is the one I have used, but has been discontinued (it's been quite a while). Anyway these little devices hold a plunger in place with a permanent magnet, but release the plunger when a coil is energised that counters the magnet. The holding force is not great, but more than enough to keep the safety on a mechanical trigger. The whole device can be potted and sealed (ideally under vacuum to eliminate voids). When pushing the plunger in to arm the solenoid, there is a definite click when the magnet kicks in, to confirm the locked state.
A similar device is the electropermanent magnet, which doesn't have a plunger, in fact it has no moving parts. You provide the steel piece that that this device will release when energised, as with a latching solenoid. It generally has greater holding force than a latching solenoid. I've used these in a seawater application. It's worth noting that there exist ferromagnetic stainless steels that can be used here to avoid corrosion.
Thanks,
-harold
Thermal cameras for monitoring visitors in highly vulnerable conservation areas
21 June 2022 3:44pm
22 September 2023 12:00pm
I would be interested how you setted up this system. Which model do yu use and how are they connected?
6 October 2023 12:27pm
We have the the FLIR FC series (FC 618) thermal cameras setup. with regards to the connection between the cameras and monitoring station, this through fiber cable and microwave radio links.
15 December 2023 8:09pm
Hi,
This is a really late answer but I was new to wildlabs then. I have a security appliance that uses state of the AI models and user defined polygon areas of interest that generates video alerts of intrusions in typically under a second.
Although its setup to install automatically on an NVidia AI on the edge boxes of your intentions were to monitor a great deal of cameras you could also install it on a desktop with a high end GPU for very high performance. At home I use a desktop with an rtx 2080ti and monitor around 15 cameras and a thermal imaging camera (old one).
I have also tested a high end model (yolov7) on a high end thermal imaging camera image and it works fine as well.
Thermal imaging cameras are hellishly expensive though and I’ve found that new extremely light sensitive cameras like the HIKvision colorvu series almost obsoletes them in terms of people detection at night at a fraction of the cost.
If you are interested I’d be happy to show you a demo in a video meeting sometime if you like. I’m pretty sure it would meeting all your intrusion detection and alerting needs.
My project page is
Video camera trap analysis help
21 November 2023 7:49am
14 December 2023 10:52am
Awesome! Will look into it.
14 December 2023 10:52am
Thanks so much, will look into it.
15 December 2023 4:48pm
Hi there!,
You should definitely check out VIAME, which includes a video annotation tool in addition to deep learning neural network training and deployment. It has a user friendly interface, has a publicly available server option that mitigates the need for GPU enabled computer for network training, and has an amazing support staff that help you with your questions. You can also download the VIAME software for local use. The tool was originally developed for marine life annotation, but can be used for any type of video or annotation (we are using it to annotate pollinators in video). Super easy to annotate as well. Worth checking out!
Cheers,
Liz Ferguson
Underwater camera trap for amphibian - Call for early users
13 December 2023 9:57pm
Hi!
CAMPHIBIAN project develops an underwater camera trap for amphibian. The interest has been proven on newts and grass snake in NW Europe. We are looking for early users worldwide, diversify target species and habitat types.
See link for more info, thanks for sharing!
Xavier
Shedding light on nocturnal behaviour: A cost-effective solution for remote, infrared video recording in the field
13 December 2023 5:55pm
Project support officer - Conservation Tech
11 December 2023 10:24pm
Is anyone or platform supporting ML for camera trap video processing (id-ing jaguar)?
27 November 2023 10:49am
7 December 2023 12:42pm
Hey there community! Im new here and looking after lots of answers too! ;-)
We are searching aswell for the most ideal App / AI technology to ID different cats, but also other mammals if possible
- Panthera onca
- Leopardus wiedii
- Leopardus pardalis
and if possible:
- Puma concolor
- Puma yagouaroundi
- Leopardus colocolo
- Tapirus terrestris
Every recommendation is very welcome, thanks!
Sam
7 December 2023 2:31pm
Is this from camera trap images? Videos? Wildlife Insights is great! It uses Megadetector as a first pass to filter out empty images (those without animals), which you can also just use on its own. If you want to do individual re-ID, you can use Wildme. And as I mentioned previously in the comments, Zamba Cloud can work with videos. Conservation AI is another player in the game as well.
11 December 2023 1:38pm
Also, take a look at TrapTagger. It has integration with WildMe.
Automatic extraction of temperature/moon phase from camera trap video
29 November 2023 1:15pm
1 December 2023 2:35pm
Hi Lucy
As others have mentioned, camera trap temperature readouts are inaccurate, and you have the additional problem that the camera's temperature can rise 10C if the sun shines on it.
I would also agree with the suggestion of getting the moon phase data off the internet.
1 December 2023 2:38pm
Do you need to do this for just one project? And do you use the same camera make/model for every deployment? Or at least a finite number of camera makes/models? If the number of camera makes/models you need to worry about is finite, even if it's large, I wouldn't try to solve this for the general case, I would just hard-code the pixel ranges where the temperature/moon information appears in each camera model, so you can crop out the relevant pixels without any fancy processing. From there it won't be trivial, exactly, but you won't need AI.
You may need separate pixel ranges for night/day images for each camera; I've seen cameras that capture video with different aspect ratios at night/day (or, more specifically, different aspect ratios for with-flash and no-flash images). If you need to determine whether an image is grayscale/color (i.e., flash/no-flash), I have a simple heuristic function for this that works pretty well.
Assuming you can manually define the relevant pixel ranges, which should just take a few minutes if it's less than a few dozen camera models, I would extract the first frame of each video to an image, then crop out the temperature/moon pixels.
Once you've cropped out the temperature/moon information, for the temperature, I would recommend using PyTesseract (an OCR library) to read the characters. For the moon information... I would either have a small library of images for all the possible moon phases for each model, and match new images against those, or maybe - depending on the exact style they use - you could just, e.g., count the total number of white/dark pixels in that cropped moon image, and have a table that maps "percentage of white pixels" to a moon phase. For all the cameras I've seen with a moon phase icon, this would work fine, and would be less work than a template matching approach.
FYI I recently wrote a function to do datetime extraction from camera trap images (it would work for video frames too), but there I was trying to handle the general case where I couldn't hard-code a pixel range. That task was both easier and harder than what you're doing here: harder because I was trying to make it work for future, unknown cameras, but easier because datetimes are relatively predictable strings, so you know when you find one, compared to, e.g., moon phase icons.
In fact maybe - as others have suggested - extracting the moon phase from pixels is unnecessary if you can extract datetimes (either from pixels or from metadata, if your metadata is reliable).
5 December 2023 10:09pm
camtrapR has a function that does what you want. i have not used it myself but it seems straightforward to use and it can run across directories of images:
https://jniedballa.github.io/camtrapR/reference/OCRdataFields.html
Paving the Way for Women: LoRaWAN Technology in Akagera National Park with Clementine Uwamahoro
29 November 2023 5:22pm
ICOTEQ launch TAGRANGER® system of products
23 November 2023 1:25pm
AWMS Conference 2023
Eliminatha, WiCT 2023 Tanzania
21 November 2023 1:09pm
Passionate wildlife researcher and tech user, making strides in Grumeti, the heart of western Serengeti,Tanzania, using Camera Traps to gain priceless insights into the lives of this unique fauna and contributing greatly to understanding and preserving the Serengeti's ecosystems.
setting up a network of cameras connected to a server via WIFI
3 April 2022 7:19am
9 July 2022 6:03am
Great discussion! Pet (and other 'home') cams are an interesting option as @antonab mentioned. I've been testing one at home that physically tracks moving objects (and does a pretty good job of it), connects to my home network and can be live previewed, all for AUD69 (I bought it on special. Normal retail is AUD80):
On the Wifi front, and a bit of a tangent, has anyone done any work using 'HaLow' (see below for example) as it seems like an interesting way to extend Wifi networks?
17 July 2022 8:21am
Cool thread!
I will be testing Reolink Wi-Fi cameras in combination with solar powered TP-Link long range Wi-Fi antennas/repeaters later this field season for monitoring arctic fox dens at our remote off grid site in Greenland. The long range Wi-Fi antennas are rather power hungry but with sufficient solar panel and battery capacity I am hopeful it will work.
I am looking forward to explore the links and hints above after the field season.
Cheers,
20 November 2023 9:39am
Thank you for the links, Robin.
Alternative to Reconyx Ultrafire
8 November 2023 9:07am
18 November 2023 8:22pm
The two cameras you mention below tick off most of the items in your requirements list. I think the exception is the “timed start” whereby the camera would “wake up” to arm itself after a certain date. Camlockbox.com provides security boxes for both.
Especially if a white flash is useful in your research, you may also want to consider the GardePro T5WF. I don’t have a lot of long-term experience with this camera, but it is one of the few that offers a white flash, and it has excellent battery life, especially for night captures. The audio can be a little flaky
I have done posts on these cameras, including a teardown. See:
https://winterberrywildlife.ouroneacrefarm.com/2022/04/10/browning-spec-ops-elite-hp5-teardown/
https://winterberrywildlife.ouroneacrefarm.com/2022/09/26/inside-the-bushnell-core-ds-4k-trail-camera/
https://winterberrywildlife.ouroneacrefarm.com/2023/11/18/gardepro-t5wf-white-flash-trail-camera/
I have heard reports that the HP5 can let in moisture in very wet environments. This may be a direct water contact type of thing, as we have never had water issues with this camera when it is installed in a lock box (US Northeast, Northwest).
We prefer the HP5 due to superior image and audio quality. That said, there is a known issue that with some HP5 cameras, with some fast (> 80 MB/s rated read) and large SD cards, the SD card can become corrupted, preventing the camera from capturing images. I address this, including a fix via firmware, in another post:
https://winterberrywildlife.ouroneacrefarm.com/2023/11/16/fixing-browning-edge-elite-hp4-and-hp5-sd-card-corruption/
Hope this helps.
-bob
Ideas for easy/fast maintenance of arboreal camera traps
30 August 2023 8:51pm
15 November 2023 9:04pm
I use the same wifi trick with Reolink solar cameras looking at tree cavities (Austrian mountain forest). You can even put the mobile router on drone to get connection to the cameras.
16 November 2023 10:29pm
Yup Reolink is awesome! I've had a few cameras die on me, but hopefully I can get them repaired.
17 November 2023 9:19am
Hi Ben,
I would be interested to see if the Instant Detect 2.0 camera system might be useful for this.
The cameras can transmit thumbnails of the captured images using LoRa radio to a Base Station. You could then see all the captured images at this Base Station, as well as the camera's battery and memory information (device health). In addition, you could also change camera settings from the Base Station so you would not need to reclimb the trees to change from PIR sensitivity high to medium for instance.
The Instant Detect 2.0 cameras also have an external power port so a cable could be run to the ground to a DC 12V battery for long term power.
If you wanted to, you could also connect the Base Station to the Cloud using satellite connectivity, so that you can monitor the whole system remotely, but this will incur more cost and power usage of the Base Station.
I'd be keen to hear your thoughts,
Thanks,
Sam
Instant Detect 2.0 and related cost
16 November 2023 12:50am
16 November 2023 10:55am
Hi Kaarthika, hi all,
ZSL's Instant Detect 2.0 is currently undergoing Beta testing with external partners and so is still pre-production. We therefore do not have final pricing for the system.
Saying this, we have got a manufacturing partner fully set-up who has already completed two full build rounds of the system, one in 2020 and another in 2023. This means we actually have a very good idea of the system's build costs and what these are likely to be when we can manufacture the system in volume.
While I cannot release this pricing yet, I am confident that we will have an unparalleled proposition.
In particular, the satellite airtime package we can supply to conservationists due to the generosity of the Iridium Satellite Company means that each system can send 3,600 (25-50KB) images a month from anywhere in the world for a single fixed fee. This equates to around a 97% discount to the normal commercial rates.
We are currently very busy fundraising so that we can make this final step to scale the system.
If we can secure this funding, we hope to go into volume production by mid-2024.
Best wishes,
Sam
16 November 2023 2:12pm
Thank You for your valuable update Sam
16 November 2023 2:30pm
Hello Sam,
What would you say would be the estimate cost was for the first version Instant Detect 1.0 ? That might help my research ?
Research/Field Technician - SPEC Lab Ocelot Research Program
16 November 2023 1:58pm
Insight; a secure online platform designed for sharing experiences of conservation tool use.
7 November 2023 1:01pm
A secure platform designed for those working to monitor & protect natural resources. Insight facilitates sharing experience, knowledge & tools to increase efficiency & effectiveness in conservation. By sharing we reduce time & money spent to find, test, & implement solutions.
DeepFaune: a software for AI-based identification of mammals in camera-trap pictures and videos
14 July 2023 3:14pm
24 October 2023 8:46pm
Hello to all, new to this group. This is very exciting technology. can it work for ID of individual animals? we are interested in Ai for identifying individual jaguars (spots) and andean Bears (face characteristics). Any recommendation? contact? thanks!
German
25 October 2023 8:57am
That's a very interesting question and use case (I'm not from deepfaune). I'm playing with this at the moment and intend to integrate it into my other security software that can capture and send video alerts. I should have this working within a few weeks I think.
The structure of that software is that it is two stage, the first stage identifies that there is an animal and it's bounding box and then there's a classification stage. I intend to merge the two stages so that it behaves like a yolo model so that the output is bounding boxes as well as what type of animal it is.
However, my security software can cascade models. So if you were able to train a single stage classifier that identifies your particular bears, then you could cascade all of these models in my software to generate an alert with a video saying which bear it was.
4 November 2023 4:51am
Hi @GermanFore ,
I work with the BearID Project on individual identification of brown bears from faces. More recently we worked on face detection across all bear species and ran some tests with identifying Andean bears. You can find details in the paper I linked below. We plan to do more work with Andean bears in 2024.
I would love to connect with you. I'll send you a message with my email address.
Regards,
Ed
Nepal's tiger conservation gets tech boost with AI-powered deer tracking
30 October 2023 1:23pm
Researchers in Nepal are using vertical cameras and AI technology to track and profile individual spotted deer (Axis axis), similar to the methods used for tigers.
Modifying GoPro cameras to be IR sensitive.
25 October 2023 6:38pm
27 October 2023 6:35am
Hi Jay!
Thanks for posting this here as well as your great presentation in the Variety Hour the other day!
Cheers!
17 January 2024 2:54am
Hi Dhanu,
Our group moved to Wildlife Insights a few years back (for a few reasons but mostly ease of data upload/annotation by multiple users) so I haven't tried EcoAssist. This being said, I will look into it as a pre-WildlifeInsights filter to analyze the tens of thousands of images that get recorded when camera traps start to fail, or get confused with sun spots (which can be common at one of our sites, a south-facing slope with sparse canopy cover).
Thanks for sharing!