Wildlife Population Monitoring

Bioacoustics can be used to protect big cats. A leopard’s roar, a deep, rasping call evocative of a saw slicing through wood, may hold the key to tracking these elusive felines. In a study published in Remote Sensing in Ecology and Conservation, researchers found that bioacoustics—a method of analyzing animal sounds—could enable more efficient leopard monitoring across vast landscapes, reports DANN OKOTH. The study, conducted in Tanzania’s Nyerere National Park, paired autonomous recordings with camera traps to analyze the vocalizations of large carnivores. The researchers identified individual leopards with an accuracy of up to 93%, suggesting that their “sawing” roars are as unique as fingerprints. Jonathan Growcott, a Ph.D. student at the University of Exeter and one of the study’s authors, called the findings “an important first step towards using bioacoustics in the conservation of leopards.” The implications could be significant. Leopards, classified as vulnerable by the IUCN Red List, face mounting threats from habitat loss and human-wildlife conflict. Traditional monitoring relies on camera traps, which are costly and difficult to deploy in remote areas. Acoustic monitoring, by contrast, offers a non-intrusive alternative. Reliable population monitoring is an essential component of any wildlife conservation program, said Paolo Strampelli. With further refinement, eavesdropping on leopards may become a powerful conservation tool. 🐆 Leopards’ unique ‘sawing’ sounds could aid conservation efforts: https://lnkd.in/gPQcTrUf

What if conservationists didn’t have to wait months to know what’s happening in the wild? That’s the promise of Sentinel, an AI-powered device developed by the nonprofit Conservation X Labs. Unlike traditional camera traps that sit passively for months before data is retrieved, Sentinel acts as a mini-computer in the field. It processes images locally, filters out empty frames, identifies species, and transmits compressed insights in real time — even from remote forests and islands with limited connectivity. Already, the technology is proving game-changing: In New Zealand, Sentinel helped detect invasive rats on Ulva Island, enabling fast action that prevented ecosystem collapse. In the Florida Everglades, it has been deployed to track invasive Burmese pythons. Across 14 countries, more than 500 units are now in use for invasive species detection, anti-poaching efforts, and wildlife monitoring. Challenges remain — from training AI models on rare species with limited images to keeping costs low enough for underfunded conservation groups. But the potential is huge: turning camera traps from passive observers into real-time guardians of biodiversity. 👉 Read Abhishyant Kidangoor's interview with Dante Wasmuht, head of AI at Conservation X Labs, on how Sentinel works and why scaling it could transform conservation: https://lnkd.in/gtAY_qUu. Image: With the use of Sentinels, organizations around the world are attempting to reduce the time between camera trap monitoring and conservation action on the ground. Image courtesy of Conservation X Labs.

NEW PAPER - CAMERA TRAPS. A bit late to post, but proud to have contributed to a major new paper—over 100 field biologists from across Australia came together to analyse a decade of camera trap use. It’s the largest effort of its kind, and the findings are clear. Despite widespread adoption, camera trap efforts have plateaued, analyses remain basic, and data sharing is limited. Our review highlights the need for a paradigm shift. We propose the Wildlife Observatory of Australia (WildObs) to: • Standardise metadata and image sharing • Harness AI/ML to ease the ID bottleneck • Support hierarchical modelling & big data analytics • Build capacity and foster collaboration Australia has the tools and talent to lead the world in wildlife monitoring—if we embrace smarter, connected science. 🌏📸 https://lnkd.in/gbtw8MvV #CameraTraps #WildObs #StantecProud #StantecScientist #ConservationScience #BigData #Ecology #AIforWildlife

Animals across the globe are running out of places to live. Populations of vertebrates such as monkeys, fish and birds declined 73% on average between 1970 and 2020, according to the World Wildlife Fund. This post, bylined by yours truly in collaboration with Matt Ball, details how the Map of Life Rapid Assessments (MOLRA) system uses drones, sophisticated sensors, and AI to collect audio, visual, and genetic samples to support detailed species mapping—where plants and animals are likely to be and how those populations are changing.   Together with partners and support from Sony, Esri, the Field Museum, Rutgers University, the Federal University of Amazonas, and the E.O. Wilson Biodiversity Foundation, the Yale University Map of Life (MOL) team developed MOLRA for last year’s biodiversity XPRIZE Rainforest competition. The runner-up prize award of $2 million is now being used to fund a spin-off company, Map of Life Solutions.   “As pressures on lands are increasing and climate change accelerates, it is ever more important that we are guided by evidence and quantitative metrics to achieve positive outcomes for biodiversity,” said Walter Jetz, professor of ecology and evolutionary biology at Yale, who also leads the Map of Life project. “And it is critical that we ensure this data reaches the desks of decision-makers and is informed by their needs. This is why we are so excited about Map of Life Solutions offering the full link from new data collection technologies to the custom reporting that organizations require.”   https://lnkd.in/gWj4H9Bx

How to Use Sound and AI to Protect the Environment https://lnkd.in/giquV4Ai If you haven’t been down a similar rabbit hole, you may not have heard of Rainforest Connection (RFCx) , a nonprofit that employs AI to help preserve the world’s rainforests. Rainforest Connection’s core idea is simple but ingenious: listen (literally) to the rainforest. The organization collects sound from rainforests in 37 countries across the globe, recording audio from what equates to nearly 2 million acres of land, much of it in remote corners of dense jungle. Rainforest Connection’s AI tools parse this data to do things like monitor biodiversity and detect instances of illegal deforestation or animal poaching. But their work doesn’t just take place behind a computer screen; Rainforest Connection collaborates with local organizations and Indigenous communities to guide conservation efforts and facilitate real-time interventions to stop illegal activities. Natasha Uzcátegui-Liggett: Can you tell me a little bit more about the work that Rainforest Connection does? Bourhan Yassin: Rainforest Connection uses sound as a way to understand what’s going on in forests and other ecosystems. Sound is a really good way of being able to detect things from a long distance, and it works in a 360-degree fashion. We use sound in a couple of different ways. One, we use sound to detect illegal activities. Sound in combination with AI can detect things like chainsaws, which are indicative of illegal logging, or gunshots, which are indicative of poaching. We also use sound to do what’s called “biodiversity monitoring.” We’re able to use AI to detect individual types of species and even the type of song or vocalization. A lot of indicator species, species that indicate the health of the forest, are vocal, like birds and amphibians, so monitoring their sounds also allows us to measure the health of the forest. The success of many projects, whether it’s a restoration project or a project to create a protected wildlife area, depends on showing whether biodiversity is improving because, ultimately, that’s the essence of what you’re trying to do. Using a combination of AI, machine learning, and a bunch of tools that we’ve developed, we’re able to use sound to track this. For most people in the scientific community, it seems like publishing a paper is their sort of nirvana moment, but translating that into conservation action is often difficult. The idea behind Arbimon was to provide tools for scientists to fast-track their research, allowing them to get to the point of making discoveries more quickly while also providing a tool for on-the-ground conservation organizations, enabling them to decipher the scientific findings pretty easily.

The Conservation Superpowers of IoT... May 17th marks a crucial day for our planet. Today, it's Endangered Species Day. As we spotlight the urgent need to protect our planet's biodiversity, I want to remind you how cutting-edge technology, specifically IoT, is a game changer in wildlife conservation. Here’s why IoT isn’t just about smart homes and industries: → Real-Time Monitoring: IoT devices can track the health and movement of species across remote areas, sending alerts about potential threats—like poaching or habitat encroachment—in real-time. → Habitat Management: Automated sensors can monitor environmental changes, helping conservationists act swiftly to mitigate adverse conditions. → Data-Driven Decisions: Accumulated data from IoT devices can guide better decision-making processes in conservation strategies, ensuring efforts are directed where needed. 🌱 Want to get involved? → Support local and global conservation initiatives, adopt eco-friendly practices, and consider participating in habitat restoration projects. 💡 Let’s discuss: How can technology further revolutionize our approach to saving endangered species? 🔄 Share if you believe in tech-powered conservation! #EndangeredSpeciesDay #IoT #ConservationTech #Biodiversity #EcoInnovation

Did you know drones can track wildlife using animals’ thermal signatures? A recent study in Queensland, Australia processed thermal imagery collected by drones with machine learning algorithms designed to detect koalas’ specific heat signatures compared to other animals. Additionally, the koala population sampled already had radio collars, so the thermal identification results could be compared to radio detection results to assess how well the method performed. Researchers found that the heat signature model could detect about 86% of the individuals present. In comparison, expert spotters can only detect about 70% of koalas on average. Conservation AI produces AI models for animal detection in a variety of habitats, such as savannah or underwater marine ecosystems. Other projects have included species including spider monkeys, elephants, clownfish, and Arabian oryxes. However, most of these projects use simple visual identification rather than infrared signature recognition. The Queensland koala project is notable because it combines thermal signatures with AI models––koalas often live in complex canopy cover and are much more difficult to identify visually. While these drone-based methods are cheaper and faster than human spotters or dogs, researchers caution that this is not the only solution. Instead, thermal signatures and machine learning are useful tools for conservationists, particularly in more challenging ecosystems. Read about ConservationAI’s models: https://lnkd.in/esad4ipt Read more about the koala identification project: https://lnkd.in/eEn9afKd #fourhundredfeet #dronesforgood #drones4good #drones #wildlife #conservation #machinelearning #AI Follow me (Kat James) and to learn more about #dronesforgood. This month I'm talking about drones for wildlife conservation.

📢FineTuning RetinaNet for Wildlife Detection with PyTorch Wildlife populations are rapidly declining due to poaching, habitat loss, and climate change—traditional monitoring methods simply aren't fast or efficient enough. To tackle this, we've fine-tuned RetinaNet, a powerful deep learning model, using PyTorch, achieving an impressive 79% mAP in wildlife animal detection tasks. Our step-by-step tutorial guides you through dataset preparation, RetinaNet customization, training strategies, and optimization techniques. We've also benchmarked RetinaNet against YOLO11, demonstrating clear improvements in detection accuracy and reliability, even with limited hardware resources. Enhance wildlife conservation efforts with deep learning. Explore our RetinaNet fine-tuning tutorial and download the complete code now! Link to the article: https://buff.ly/CKb8T88

🌿 5 Must-Have Offline Apps for Forest Researchers & Wildlife Fieldwork 🐾📱 Working deep in the forest without network access? No problem. Here are 5 reliable mobile applications that work without internet and are perfect for researchers collecting data in remote forested landscapes: 📲 1. KoboToolbox Purpose: Survey design & data collection with GPS, images, and offline sync. Best For: Vegetation sampling, HWC data, socio-ecological surveys. https://lnkd.in/g_KDvQUD 🧭 2. Tracklia Purpose: GPS tracks, waypoints, and path monitoring. Best For: Mapping animal movement, patrolling paths, or vegetation transects. https://tracklia.com 🌐 3. Mappt Purpose: Mobile GIS tool with offline maps and shapefile integration. Best For: Collecting geotagged data in rugged terrain with custom base maps. https://mappt.com 🗺️ 4. QField (by QGIS) Purpose: Carry your QGIS project into the field with full offline access. Best For: Researchers using layers, shapefiles, and complex GIS setups. https://qfield.org 🐘 5. SMART Mobile Purpose: Anti-poaching patrol and wildlife monitoring app with offline support. Best For: Camera trap tracking, patrol routes, species sightings. https://lnkd.in/gqGt_8bu 💡 Why these apps matter: They help you map, record, and analyze field data without needing network—making them perfect companions for field-based conservation science. 🔗 Save this post, try them out, and share it with your research team! 🗣️ Do you use any of these tools? Have another favorite app that works offline in the wild? Drop it in the comments! 👇 #WildlifeResearch #ForestEcology #GIS #RemoteSensing #FieldData #KoboToolbox #Tracklia #QField #Mappt #SMARTConservation #OfflineApps #ConservationTech #WildlifeMonitoring #EcologyTools

💡 New paper on how the latest AI model in #WildlifeInsights works! https://lnkd.in/grBZmASZ I'm thrilled to share that our paper on "To crop or not to crop: Comparing whole-image and cropped classification on a large dataset of camera trap images" has been published in the journal IET Computer Vision Special Issue: Camera Traps, AI & Ecology. This research represents a significant step forward of applying AI to wildlife conservation and biodiversity monitoring. We present a novel deep learning framework for accurately detecting and identifying wildlife species from camera trap images. Our approach leverages MegaDetector v5 for wildlife detection followed by species classification on a cropped image of the animal. The ability to automatically process vast amounts of camera trap data in Wildlife Insights is saving researchers hundreds of hours of time better spent doing real science instead of combing through images manually. 🦊 If you use camera traps, give the model a try yourself in https://lnkd.in/gW7Zjaap, and let us know what you think! 🦌 Thanks to all in the Acknowledgments section of the paper who helped make this come to fruition, and special thanks to the Googlers who helped over the years on model development, whether by giving their 5% time or 100% time to Wildlife Insights: Tomer G. Ștefan Istrate Dan Morris Hyungwon (Clint) Kim Sara Beery Burcu Karagol Ayan Michael Brooks Shivaji Dutta Logan Goldberg Jacqueline Hea Jonathan Huang Zachary Hynes Simon Ilyushchenko Charbel Kaed Sayali K. Chen Luo Pravar Mahajan Bryan Rutkowski Karen Y. W. Leon Zhou and with the strong executive support from Karin Tuxen-Bettman Matthew Hancher and Rebecca Moore!