Animal Detecting Week 9 - TASMANIA 🦘 I always thought of this island as of something magical. So distant, so unreachable... So I was happy to stumble upon a Tasmanian dataset I could play with. From 400 images, animals detected on 374 images (93.5%), with 417 total animals & humans detected.

You can now delete detections in a project. You asked - we added. 😎 Not all detections are always useful. Sometimes you might look for a coyote, but there are too many seagulls cluttering your data. Now you can delete them all with one click (and typing DELETE for safety!).

Our front page got an update! 🦛 Now you can test if we can recognize your animal directly on the front page. Just choose your country & drag and drop an image. PS. Photo used is by Marc Mol, taken in South Luangwa National Park, Zambia 💚

It was refreshing to attend NetworkNature EU conference in Brussels this week. Normally I communicate a lot with people working "close to the source" - wildlife organizations and agencies overseeing nature areas. So it was new for me to meet people on the other end - policymakers alike.

Ahh, now I have something to wear! 🤠 Our new merch arrived, time to detect some animals!

Animal Detecting Week 8 🐾 European Mammals & birds with LoFi sounds for a sunny Friday. 🌆 After working with many African and American datasets I felt curious about how we would perform on home datasets - in Europe. So I run Animal Detect through this German dataset and the results are amazing.

Camera trap & drone image analysis has never been easier. Really. In Animal Detect we listen to your feedback very carefully! What you are hyped about, what you think is "meh", and everything in between. We have just upgraded species labeling experience to a new level. Now you can:

Animal Detect is free for students! Writing a paper, doing your PhD, or a regular semester project? We are here to help. As recent ex-students we remember how scrappy the times get. You are stressed about 1000 and 1 things in your life. The technical work might be your Priority 8.

Animal Detecting Week 7 We got leopards in Botswana today 🐆 Thread 🧵👇

Animal Detecting Week 6 🦓 𝟏𝟎𝐆𝐁 𝐨𝐟 𝐢𝐦𝐚𝐠𝐞𝐬, 𝟐𝟎𝟎𝟎𝟎+ 𝐚𝐧𝐢𝐦𝐚𝐥𝐬. 𝐀𝐋𝐋 𝐔𝐍𝐃𝐄𝐑 𝟒𝟓 𝐌𝐈𝐍𝐔𝐓𝐄𝐒 We updated both processing steps in our system - Filtering blank images and classifying animals. Both steps got a whooping 400% and 500% speed boost.

Animal Detecting Week 5 🐾 𝐁𝐄𝐖𝐀𝐑𝐄, 𝐋𝐈𝐎𝐍𝐒, 𝐑𝐇𝐈𝐍𝐎𝐒, 𝐚𝐧𝐝 𝐦𝐚𝐧𝐲 𝐨𝐭𝐡𝐞𝐫𝐬! For this week we used "Desert Lion Conservation Camera Traps" from Northern Namibia. Big shoutout to Desert Lion Conservation for their conservation efforts towards desert lions.

Animal Detecting Week 4 For this week цу used Wellington Camera Trap dataset from New Zealand. Lets see! Dataset Details Images: 1000 (slice) Size: 729MB

Animal Detect API Service is live! We are releasing our endpoint /predict Send it an image -> Receive all detections with animal species, for your country. The setup takes 1 minute. Try it out now for free and give us feedback. Link to start page: www.animaldetect.com/api

Hey everyone! 😄 We made a free online tool to analyze single images for wildlife community. It works extremely simple - drop an image, get animal (or human/vehicles) detected/labelled in seconds.

Animal Detect can help your wildlife organization to build cutting-edge data & AI driven solutions! If you drown in wildlife data, frustrated with old software that doesn't feel good, or dream of an AI model trained specifically for your species: ❗ we can help. read further what can we do 👇

Classification just received a MASSIVE boost in performance. It is literally 5 times faster now. Processing 1000 images / 1GB used to take 10+ minutes. Now its under 2 minutes. We are on a hunt to customize and improve every single step of the process on Animal Detect.

Animal Detecting Week 3 In this weekly post I am processing various animal datasets and report on results. Sometimes those datasets are purely wildlife, sometimes they are mixed with human-made objects and AI generated content. Lets see!

You can not filter your projects by animals that were detected in them. Thanks for your feedback and please continue writing it! 🐺 Animal Detect.

You can now select and relabel many detections at once in Animal Detect. - Hold SHIFT + Click on image to select 1 image at a time - CTRL (or ⌘) + A to select all images in the current view PS. all the cats on the video were not really classified as 'goats' by AI, it is only for demonstration. :^)

I love what I do. Do you?

Animal Detecting Week 2 In this weekly post I am processing various animal datasets and report on results. Sometimes those datasets are purely wildlife, sometimes they are mixed with human-made objects and AI generated content. Lets see! Dataset: Outskirts of USA Images: 500

Heya everyone! The platform received an update that should make it easier to navigate through pages, and use less resources, both internet traffic and RAM. See ya later, alligator. 🐊

Took our caticus Yuki outside for the first time. The little guy was scared at first and did not dare to leave his transportable bag for 5 minutes. But then he did! And he loved it. All for you my little one. 🥰

This is so bizarre I can't believe it still exists. 🤯 It is legal in Denmark to drive/park your car on the sea shore. It feels absolutely outlandish and barbarian towards nature.

We have completely reworked Animal Detect experience. Now you can experience fully automated classification of your animal species! The platform is now smarter, faster and much easier to use. Try it and give us your feedback! Here is a short demo using the new page: www.youtube.com/watch?v=TP-E...

I am not afraid or shy to say that we are about to drop THE BEST animal classification tool. We already have it. It's all about building it into Animal Detect platform! Working extra hard to get it done this month. Happy World Environment Day. 🌎 Here is a photo of my catsy reading a book:

Animal Detect got a fresh homepage look! 🙈 Now more information and clarity about what can we do! Missing something? Let me know. ✍️

Can Animal Detect support underwater monitoring & fish data processing? Yes, it can. Marine people experience the same stuggles as their on-land colleagues: 𝗖𝗼𝗹𝗹𝗲𝗰𝘁𝗲𝗱 𝗳𝗶𝘀𝗵 𝗶𝗺𝗮𝗴𝗲𝘀 -> 𝗻𝗲𝗲𝗱 𝘁𝗼 𝗽𝗿𝗼𝗰𝗲𝘀𝘀 𝗶𝘁 -> "𝘂𝗴𝗵, 𝗶𝗺 𝗴𝗼𝗶𝗻𝗴 𝘁𝗼 𝗰𝗿𝘆" There is easier way, I promise: lnkd.in/d_xg5FtA 😊

User clarity is #1 priority 🐨 We are adding new pop-ups like this one to help you understand how to use Animal Detect the most efficient way! Try it out now: animaldetect.com

We are live! 🚀 Animal Detect is now open to everyone. Built for ecologists and wildlife researchers we help you to process your wildlife images EASY. 1. Upload your images 2. Use AI to filter, label, and sort 3. Download results Sign up now: app.animaldetect.com/signin

Nothing clears your brain more than a 10km hike in the sunny nature. Touching grass + touching sheep combo. Animals detected that day: 15 🐨 Panoramaruten Mariagerfjord.

Hold on team, I need to stop development. My cat is trying... to flood us? 😅 Why do cats do this?

To hell with vibe coding, Trump, tariffs, microsoft cursor attack and everything else hot this week. Hike is all you need to ventilate the brainrot accumulated in you.

Seeing new AI mega models being dropped on us every week feels like pixel/RAM race in previous decades. Except this time, every new release can completely erase whole industries or job titles. What happened to Junior Fr

take your brake RIGHT MEOW okay, okay Yuki...

Who needs users when you can focus on perfecting phone view for a desktop app? 😄

We added a new feature to Animal Detect - interactive map based on GPS tags from the uploaded images. Users can click on each animal detection pin and see the image. This will help large areas to track their animals and prepare report results. The maps can also be shared with others with a link.

What does it remind me of... 🤔

Work From Home got benefits. The benefit:

You can now generate beautiful pie charts for your projects using Animal Detect. Share them with your colleagues, stakeholders, and viewers!

Something I will proudly tell my grandkids. <Your old champ was once climbing trees and stalking elephants in Asia>

Are you BEFORE or AFTER? 🤓 By combining SpeciesNet with Animal Detect own method, we are able to classify 30% more animals on top of an already state-of-the-art SpeciesNet model. Working full-speed now to launch this to the platform and let you try it for yourselves. animaldetect.com

We have concluded two-week test phase of animaldetect.com A lot of good feedback, bugs, and improvement directions were given. We are happy with the outcome. 🥳 Real user feedback is #1 priority for us and is detrimental to our work. Now our plan is to prepare for the launch in April. Stay tuned!

This week Animal Detect went into TEST-MODE with 25 volunteers trying it! 🫣 We have received a lot of good points for improvement, and also found many critical issues ☠️ . Well, that is what testing is for, right? Here is our new video showing new UI & Features. www.youtube.com/watch?v=Dgye...

❗ 3 MORE SPOTS TO animaldetect.com CLOSED-TEST PHASE On Monday, we are starting the test phase of our first MVP version! We are looking for a limited number of testers and still have 3 places left. Just write to me if you would like a spot reserved for you. Only your email is needed.

My free app deepfocused.me was used by 162 unique visitors last week. User statistics: ☎️ 45% of visitors using phone 🌍 from 32 countries 💬 from 8 sources (LinkedIn, Facebook, Forums, Google) 🤡 1 visitor came from a paid ad... wait what? I didn't do any paid ads 😅 Try it out and give me feedback 🫣

Are you bored? I've made a simple app that generates your spirit animal in 5 seconds (free/no sign-in required) deepfocused.me I did it to promote awareness of different animal species and it’s been a great fun and exercise for my web dev skills. POST YOUR SPIRIT ANIMALS HERE, for fun. :^)

Thinking about making a sticker pack out of my glorious cats for Signal, Messenger, and WhatsApp. Should I? 😀

The impact of 'exporting extinction' to developing nations is significantly greater than the biodiversity loss within these wealthy nations' own borders, with the US alone accounting for 3% of global forest habitat destruction outside its territory. www.theguardian.com/environment/...

My partner in crime Hugo Markoff wrote a brilliant in-depth guide on how to develop your own wildlife camera, based on our 2 years experience of building one in our Animal Detect startup. It covers everything from design & hardware to implementing AI on cloud! 🤯 www.linkedin.com/pulse/buildi...