Since we launched our AI Vectorizer for #QGIS last Friday, we've served over 10,000 autocompletion requests from 80 different countries 😎 Since how to use the plugin has changed, we wanted to show the tips and tricks our users learned to digitize 2-3x faster than by hand. Here, Michael Egan walks through how to get started with the AI and get the best performance out of it:
Bunting Labs (YC S22)
Data Infrastructure and Analytics
Bunting Labs is no-code GIS infrastructure. Backed by YC (S22), founded by MIT and USC grads.
About us
Bunting Labs automates the digitization of maps into GIS and CAD. Backed by YC (S22), founded by MIT and USC grads.
- Website
-
https://buntinglabs.com/
External link for Bunting Labs (YC S22)
- Industry
- Data Infrastructure and Analytics
- Company size
- 2-10 employees
- Headquarters
- San Francisco
- Type
- Privately Held
- Founded
- 2022
- Specialties
- GIS
Locations
-
Primary
San Francisco, US
Employees at Bunting Labs (YC S22)
Updates
-
Bunting Labs (YC S22) reposted this
Je suis ravie de partager ma récente découverte ici sur LinkedIn : l'extension « Bunting Labs AI Vectorizer » développée par Bunting Labs (YC S22). Cette extension innovante facilite grandement la vectorisation de cartes grâce à l'intelligence artificielle. 🎯 Mes remarques : ✅ Les résultats ont été globalement satisfaisants. 🌟 Dans certains cas, le plugin a fonctionné bien, offrant une vectorisation presque automatique et d'une grande précision. 🤔 Cependant, dans d’autres cas, il a rencontré quelques difficultés, notamment avec les changements de direction et certaines entités complexes. Dans l'ensemble, je suis très impressionnée par les capacités de ce plugin et je recommande vivement à tous ceux qui travaillent avec QGIS de l'essayer. 📹 Voici une vidéo que j'ai réalisée pour partager mon expérience avec ce plugin. https://lnkd.in/d9jd5htF Un grand merci à l'équipe de Bunting Labs pour cette innovation ! 🚀 N'hésitez pas à partager vos expériences et vos retours si vous avez également essayé cette extension. #GIS #QGIS #Vectorization #ArtificialIntelligence #BuntingLabs #Innovation #Geospatial
Tuto QGIS: Vectorization with Bunting Labs AI Vectorizer
https://www.youtube.com/
-
Bunting Labs (YC S22) reposted this
2,960 AI experiments later, our #QGIS AI Vectorizer plugin is 3x faster than before. How’d we do it? Brendan Ashworth has created an AI that can see all possible vectorizations of your map 🗺️. By training on every style of map we could find for 50+ days on GPUs, our AI Vectorizer now knows where your vectorization is going before you do. The biggest change is your cursor location now shows the AI what you want to digitize. Because the plugin has vectorized all possible paths, clicking is more like intelligent snapping to the feature. This radical redesign fixes all of the problems of the last version: users no longer need to correct the AI’s mistakes before continuing, it has been trained extensively on aerial imagery and geologic maps, and we have increased speed by deploying our AI on servers across the world. To use the new plugin, place two vertices along the feature you want to digitize and then move your cursor to where you want the AI to go. This lets you pick between diverging paths and digitize up to confusing parts where you can manually intervene using the shift key. If you want to try this out, just update your plugin in the QGIS Plugins window and restart QGIS, or search Bunting Labs to download it. #Geospatial #GeoAI #GIS
-
Bunting Labs (YC S22) reposted this
I'm excited to go to Slovakia this September to listen to Brendan Ashworth speak at the QGIS User Conference 🇸🇰 If you'll be there, let me know!
🎤 Are you wondering what talks and presentations await you at the QGIS User Conference 2024? You can look forward to: -> Johannes Kröger on Delivering the perfect QGIS installation and environment to your users -> Brendan Ashworth on Everything I learned building AI models for QGIS 💬 The conference will hold over 50 talks and 15+ workshops spread across 6 rooms, covering everything you love about QGIS – from inspiring use cases and advanced workflows to in-depth sessions on new features and the latest trends in the QGIS world. 🎯 Stay updated with the continuously refreshed list of talks and workshops here: https://lnkd.in/dXEKzCrc 🐦 Don't miss out on our Early Bird registration: https://lnkd.in/dTjn2Nz4 📅 Date: 9-10 September 2024 📍 Location: Bratislava, Slovakia
-
Bunting Labs (YC S22) reposted this
Thank you to the 16,000+ downloads in #QGIS for our AI Vectorizer! We often get asked for tips on how to get the most out of the plugin, so I wanted to make a video to address what Brendan Ashworth and I think are the most important things to think about when using it: 🔎 1. Make sure that you are zoomed so that the feature you are digitizing is about 3 pixels wide 🎯 2. Make sure that your first completion is entirely on the line or polygon you want to digitize ❌ 3. If the AI made a mistake, hold down shift to cut the line and continue to hold it down to make manual completions. Let go to make more AI completions 📚 4. The top visible layer is what the AI will vectorize if your images overlap
-
Bunting Labs (YC S22) reposted this
I had an amazing conversation with Daniel O'Donohue about building custom AI models for #GIS. Here are my favorite parts of this podcast: ⏳ 4:41 Why build the MVP in QGIS? ⏳ 14:44 How does an AI Vectorizer actually work? ⏳ 19:55 What would AI georeferencing look like? ⌛ 25:40 When is a machine learning model "good enough"? ⌛35:30 What's the future of AI x desktop tools look like?
Brendan Ashworth the CTO and co-founder of BuntingLabs is integrating AI with #QGIS And in this episode of the MapScaping podcast, we are talking about Autocomplete for vectorization. Along the way Brendan will share with us why Bunting Labs chose to build this on top of QGIS, the Challenges in Map Digitization, what the development process was like and how this is different from tools like Segment Anything ( from meta ) https://lnkd.in/g7_uQ6vt
-
We're excited to be a Phase 2 winner of the US Census Bureau's Supply Chain Challenge! Our proposal of using artificial intelligence to monitor and detect changes in imagery of critical sites will hopefully allow the Census to provide more accurate and real-time warnings of supply chain disruptions. And congratulations to the other winners of the challenge! #supplychain #geoai
We are excited to announce the winners of the #StatVentures Supply Chain Challenge Phase 2! 🏅 Phase 2 winners submitted implementation roadmaps explaining how their approach would work to capture new or improved supply chain #data on manufacturing, imports and exports, transportation and delivery of goods, and more. The eight winning teams will receive monetary prizes with a total value of $350,000 and an opportunity to advance to Phase 3. Learn more: https://lnkd.in/e_udMG3s #CensusData #SupplyChainChallenge #tech #innovation #entrepreneurship
-
When you download our plugin in QGIS you'll automatically be on 1.10, or if you already installed it there's an option to update in the manage plugins window Give the new zoom a try and let us know what you think!
1 month and over 10,000 downloads in #QGIS later, we're releasing our most significant update to our AI Digitization plugin! Now, the plugin will digitize differently based on how zoomed in or out you are, so large features and smaller, more detailed features get a better balance of speed and accuracy (shown in the gif below). This dramatically improves performance on aerial imagery, computer generated maps, and large features on high resolution maps 👀 And, coming soon, is a completely new, more powerful AI model trained to preform better on aerial imagery, geology maps, and black and white maps, all while getting distracted less and making longer completions... #geospatial #gis #geoai
-
Bunting Labs (YC S22) reposted this
😳 Now THIS is amazing - automated AI map tracing in QGIS. Shout out to Michael Egan of Bunting Labs (YC S22) for building this amazing plugin (link in comments). 👌🏽 This is an absolutely perfect use of AI in geospatial. It takes a very manual and time-consuming task and passes that off to AI while still being able to have human intervention when needed. Amazing work! #gis #moderngis #geoai #geospatial #qgis #gischat #arcgis #esri #spatialanalysis #ai #geography #maps
-
We're excited to be launching our new QGIS plugin that autocompletes features you trace in QGIS! Thank you to everyone who we've worked with along the way to building this
6 months, 158,000 map training datapoints, and an entire rebuild of the AI later: today is the launch of our #QGIS AI Map Tracing Plugin! Brendan Ashworth and I have been working on a plugin that automates map digitization from raster to vector. How and why did we build this, and how can you try it out?👇 (Try the plugin for yourself by searching for "Bunting Labs AI Vectorizer" in the QGIS plugins window, or by visiting our website and signing up.) Our QGIS plugin is an autocomplete for lines and polygons that is a drop-in replacement the Add Line/Polygon Feature or Create Features tool. When you start tracing features in a raster map, the plugin identifies what you are tracing and completes the vector layer for you (and only the features you want, not everything in the document). By letting an AI autocomplete handle the digitization, you can spend less time digitizing each map. We wanted to build this after being surprised by how many GIS professionals complained of being slowed down because they were sent data as a PDF. Teams would have to either dedicate huge amounts of resources to slowly converting PDFs by hand or be stuck with just working off historical documents. We’ve found that the plugin can digitize features 4 times faster, and we hope that GIS professionals can see that improvement as well—spending less time on each map and opening up space for new projects. #GIS #geoai #geospatial