Загрузка видео...

Не удалось загрузить видео

На главную

VPS 2.0 Outperforms GPS Alone. Niantic Spatial’s Visual Positioning System (VPS) matches live camera data to a spatial model for precise 6DoF localization—indoors, outdoors, and in complex environments where GPS fails. 🔹 ~10-15cm accuracy in mapped areas 🔹 Works in GPS-denied environments 🔹 Persistent spatial anchors Explore this deep...

NianticSpatial's profile picture

Niantic Spatial 🌎

963,115 subscribers

19,782 просмотров • 5 месяцев назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

Today, Niantic Spatial introduces Scaniverse — our flagship product and the gateway to our spatial intelligence platform and Large Geospatial Model. Here’s what’s new: 🔹Scaniverse (iOS mobile + web): Capture once, generate multiple outputs—VPS maps, meshes, and Gaussian splats (Android coming soon) 🔹Collaborative mapping: Multi-user scans fused into a single, continuously improving model 🔹 On-device validation: Preview VPS coverage and test localization in real time 🔹VPS 2.0: Global-scale positioning—centimeter-accurate where mapped, reliable everywhere else (even where GPS fails) 🔹NSDK 4.0 (coming soon): A unified SDK across Unity, Swift, Android, and ROS 2 Read more: #NianticSpatial #Scaniverse #VPS #GeospatialAI #AI
1:00

Today, Niantic Spatial introduces Scaniverse — our flagship product and the gateway to our spatial intelligence platform and Large Geospatial Model. Here’s what’s new: 🔹Scaniverse (iOS mobile + web): Capture once, generate multiple outputs—VPS maps, meshes, and Gaussian splats (Android coming soon) 🔹Collaborative mapping: Multi-user scans fused into a single, continuously improving model 🔹 On-device validation: Preview VPS coverage and test localization in real time 🔹VPS 2.0: Global-scale positioning—centimeter-accurate where mapped, reliable everywhere else (even where GPS fails) 🔹NSDK 4.0 (coming soon): A unified SDK across Unity, Swift, Android, and ROS 2 Read more: #NianticSpatial #Scaniverse #VPS #GeospatialAI #AI

Niantic Spatial 🌎

120,591 просмотров • 5 месяцев назад

You've probably heard about GPS. But have you heard about VPS? It's the visual positioning tech that gives robots and AR glasses centimeter-level accuracy where GPS fails. Here's how it works: 0:00 - Intro 0:19 - How GPS Actually Works 1:28 - The Problem with GPS for AR & Robotics 2:06 - What is VPS? (Visual Positioning System) 2:38 - Use Cases: Delivery Robots & Drones 3:17 - VPS in Action: Apple Vision Pro & Meta Quest 3:53 - Scanning Your World with an iPhone (Multiset AI) 4:31 - Key Takeaways 4:48 - The Future: Spatial Intelligence
5:21

You've probably heard about GPS. But have you heard about VPS? It's the visual positioning tech that gives robots and AR glasses centimeter-level accuracy where GPS fails. Here's how it works: 0:00 - Intro 0:19 - How GPS Actually Works 1:28 - The Problem with GPS for AR & Robotics 2:06 - What is VPS? (Visual Positioning System) 2:38 - Use Cases: Delivery Robots & Drones 3:17 - VPS in Action: Apple Vision Pro & Meta Quest 3:53 - Scanning Your World with an iPhone (Multiset AI) 4:31 - Key Takeaways 4:48 - The Future: Spatial Intelligence

Bilawal Sidhu

14,757 просмотров • 8 месяцев назад

“Do you know where a sidewalk begins and ends?” — a simple question from Niantic Spatial CEO, Inhi Cho Suh, that gets to the heart of what today’s AI is missing. For robots, this isn’t theoretical. To safely navigate something as everyday as a sidewalk, they need to understand boundaries, context, and change in real time. Learn about how Niantic Spatial is building the missing piece in Physical AI. #NianticSpatial #HumanX #GeospatialAI #Robotics #AI #PhysicalAI
0:39

“Do you know where a sidewalk begins and ends?” — a simple question from Niantic Spatial CEO, Inhi Cho Suh, that gets to the heart of what today’s AI is missing. For robots, this isn’t theoretical. To safely navigate something as everyday as a sidewalk, they need to understand boundaries, context, and change in real time. Learn about how Niantic Spatial is building the missing piece in Physical AI. #NianticSpatial #HumanX #GeospatialAI #Robotics #AI #PhysicalAI

Niantic Spatial 🌎

18,656 просмотров • 5 месяцев назад

How do you go from scanning a real-world space → using it in an app, workflow, or environment? Our product team walks through Scaniverse step-by-step, from capturing spaces to generating production-ready 3D assets and enabling precise, real-world positioning. In the first video, Senior Product Manager, Bec Maxworth, walks through your entry point into Scaniverse from signing up to getting started so you can start building high-fidelity 3D models in minutes. Dive into the full walkthrough here: #NianticSpatial #Scaniverse #Reconstruction #VPS #GeospatialAI #AI #PhysicalAI
0:31

How do you go from scanning a real-world space → using it in an app, workflow, or environment? Our product team walks through Scaniverse step-by-step, from capturing spaces to generating production-ready 3D assets and enabling precise, real-world positioning. In the first video, Senior Product Manager, Bec Maxworth, walks through your entry point into Scaniverse from signing up to getting started so you can start building high-fidelity 3D models in minutes. Dive into the full walkthrough here: #NianticSpatial #Scaniverse #Reconstruction #VPS #GeospatialAI #AI #PhysicalAI

Niantic Spatial 🌎

10,739 просмотров • 5 месяцев назад

Large-scale Gaussian splats have reached a new level of realism. This is a well-known temple in Bangkok, reconstructed as a high-fidelity 3D environment from 360 captures. At this level, the boundary between video and 3D starts to disappear. But what you’re looking at is not a video. It’s a dense spatial representation of a real place, where geometry, texture, and structure are preserved and made machine-readable. This kind of 3D data can power Visual AI, Robotics navigation, VPS localization, XR experiences, world models, and next-generation spatial computing systems. Built with Over the Reality.
2:46

Large-scale Gaussian splats have reached a new level of realism. This is a well-known temple in Bangkok, reconstructed as a high-fidelity 3D environment from 360 captures. At this level, the boundary between video and 3D starts to disappear. But what you’re looking at is not a video. It’s a dense spatial representation of a real place, where geometry, texture, and structure are preserved and made machine-readable. This kind of 3D data can power Visual AI, Robotics navigation, VPS localization, XR experiences, world models, and next-generation spatial computing systems. Built with Over the Reality.

Over the Reality 🌐

347,394 просмотров • 3 месяцев назад

Two founders I could listen to for hours. In this conversation, Mike Horton from GEODNET and Nils Pihl from Auki give one of the clearest explanations of why both projects matter. GPS has two fundamental problems. It's not accurate enough and it doesn't work indoors. ▪️ RTK excels outdoors where visuals keep changing. Absolute, repeatable accuracy GPS can't match. ▪️Auki's decentralized VPS excels where RTK struggles, indoors and underground. Machines get precision without sending camera feeds to anyone's servers. That last point matters. Google, Tesla and others send those feeds back to their servers. Nils calls it exactly what it is: the greatest privacy threat we've ever faced as a civilisation. The Real World Web flips that because the data stays with the venue. ‼️ Together they form the positioning infrastructure for the physical world.
1:47

Two founders I could listen to for hours. In this conversation, Mike Horton from GEODNET and Nils Pihl from Auki give one of the clearest explanations of why both projects matter. GPS has two fundamental problems. It's not accurate enough and it doesn't work indoors. ▪️ RTK excels outdoors where visuals keep changing. Absolute, repeatable accuracy GPS can't match. ▪️Auki's decentralized VPS excels where RTK struggles, indoors and underground. Machines get precision without sending camera feeds to anyone's servers. That last point matters. Google, Tesla and others send those feeds back to their servers. Nils calls it exactly what it is: the greatest privacy threat we've ever faced as a civilisation. The Real World Web flips that because the data stays with the venue. ‼️ Together they form the positioning infrastructure for the physical world.

Krypto Insider 💫

13,371 просмотров • 1 месяц назад

Every Pokémon Go player unknowingly trained future robots. Over the years, players photographed 30 billion images from cities around the world. And now that data is teaching delivery robots to navigate streets. Niantic (company behind Pokémon Go), spun out an AI company called Niantic Spatial that's turning years of player data into a visual positioning system for machines. When 500 million people installed the game and pointed their phones at buildings, landmarks, and street corners, they were also capturing detailed location data that normal GPS can't match: > GPS signals bounce off skyscrapers and lose accuracy by up to 50 meters in dense cities. > That's a dealbreaker for sidewalk delivery robots that need to stop at exactly the right point. > Niantic Spatial trained a model on those 30 billion images to pinpoint a robot's exact location based on what its cameras see, not satellite signals. > Delivery startup Coco Robotics is the first partner, using the system to guide thousands of robots across LA, Chicago, and Miami. Don't think anyone thought they were training robots when they were chasing Pikachu during COVID.
1:17

Every Pokémon Go player unknowingly trained future robots. Over the years, players photographed 30 billion images from cities around the world. And now that data is teaching delivery robots to navigate streets. Niantic (company behind Pokémon Go), spun out an AI company called Niantic Spatial that's turning years of player data into a visual positioning system for machines. When 500 million people installed the game and pointed their phones at buildings, landmarks, and street corners, they were also capturing detailed location data that normal GPS can't match: > GPS signals bounce off skyscrapers and lose accuracy by up to 50 meters in dense cities. > That's a dealbreaker for sidewalk delivery robots that need to stop at exactly the right point. > Niantic Spatial trained a model on those 30 billion images to pinpoint a robot's exact location based on what its cameras see, not satellite signals. > Delivery startup Coco Robotics is the first partner, using the system to guide thousands of robots across LA, Chicago, and Miami. Don't think anyone thought they were training robots when they were chasing Pikachu during COVID.

Rowan Cheung

38,329 просмотров • 6 месяцев назад

Robots are gaining spatial awareness with Over the Reality. 🤖 Powered by the world’s largest 3D dataset, we enable robots to see, sense, and collaborate in real-world environments. This early test shows how a Unitree Go2 perceives its surroundings through OVER’s VPS — on the right, the real life robot; on the left, its real-time perception of the position inside one of OVER’s 3D maps. Robotic spatial awareness is here, living in 155,000+ mapped locations and expanding every day. One step closer to Physical AI. Stay tuned for our upcoming ROS 2 integration, unlocking a universal framework for robotic perception leveraging OVER's infrastructure!
0:25

Robots are gaining spatial awareness with Over the Reality. 🤖 Powered by the world’s largest 3D dataset, we enable robots to see, sense, and collaborate in real-world environments. This early test shows how a Unitree Go2 perceives its surroundings through OVER’s VPS — on the right, the real life robot; on the left, its real-time perception of the position inside one of OVER’s 3D maps. Robotic spatial awareness is here, living in 155,000+ mapped locations and expanding every day. One step closer to Physical AI. Stay tuned for our upcoming ROS 2 integration, unlocking a universal framework for robotic perception leveraging OVER's infrastructure!

Over the Reality 🌐

263,848 просмотров • 11 месяцев назад

Big news from OVER! 🚀 We're expanding our Visual Positioning System (VPS) globally integrating open datasets from Mapillary & Panoramax. Precise AR re-localization everywhere. Yes, EVERYWHERE! 🗺️ Mapillary & panoramax offer over 2 BILLION geolocalized images under CC-BY-SA license. Mapillary (recently acquired by Meta) covers the globe, while Panoramax specializes in France. These datasets beautifully complement OVER's community-driven OVRMaps. While they're less dense (about 1-2 orders sparser), their scale is massive. Check out Mapillary’s global coverage here: But leveraging such sparse, varied data isn't easy—think limited overlap, camera distortion, and no metric scaling. Yet, we've cracked the code! 💡 Our recent breakthroughs combining our VPS pipeline with cutting-edge Large Geospatial Models (LGMs) allow us to build accurate 3D digital twins and offer reliable VPS services using these sparse datasets. 📍 Accuracy might be lower than our detailed community maps, but it still blows GPS out of the water! This is HUGE for OVER. It means we’re extending our coverage WAY beyond the 110,000+ locations already mapped by our community. Web3 DePIN + Open Source beats Web2 giants like Niantic! And that's not all—we're also unlocking exciting new possibilities soon: 1️⃣ Faster and easier mapping through our Map2Earn program using 360-degree camera integration. 2️⃣ Integrating datasets from new DePIN partners—like our recent partnership with NATIX Network the OVER ecosystem. Stay tuned! The future of AR Spatial Computing and decentralized mapping is here, and it's open, collaborative, and unstoppable!
0:24

Big news from OVER! 🚀 We're expanding our Visual Positioning System (VPS) globally integrating open datasets from Mapillary & Panoramax. Precise AR re-localization everywhere. Yes, EVERYWHERE! 🗺️ Mapillary & panoramax offer over 2 BILLION geolocalized images under CC-BY-SA license. Mapillary (recently acquired by Meta) covers the globe, while Panoramax specializes in France. These datasets beautifully complement OVER's community-driven OVRMaps. While they're less dense (about 1-2 orders sparser), their scale is massive. Check out Mapillary’s global coverage here: But leveraging such sparse, varied data isn't easy—think limited overlap, camera distortion, and no metric scaling. Yet, we've cracked the code! 💡 Our recent breakthroughs combining our VPS pipeline with cutting-edge Large Geospatial Models (LGMs) allow us to build accurate 3D digital twins and offer reliable VPS services using these sparse datasets. 📍 Accuracy might be lower than our detailed community maps, but it still blows GPS out of the water! This is HUGE for OVER. It means we’re extending our coverage WAY beyond the 110,000+ locations already mapped by our community. Web3 DePIN + Open Source beats Web2 giants like Niantic! And that's not all—we're also unlocking exciting new possibilities soon: 1️⃣ Faster and easier mapping through our Map2Earn program using 360-degree camera integration. 2️⃣ Integrating datasets from new DePIN partners—like our recent partnership with NATIX Network the OVER ecosystem. Stay tuned! The future of AR Spatial Computing and decentralized mapping is here, and it's open, collaborative, and unstoppable!

Over the Reality 🌐

1,313,342 просмотров • 1 год назад

Have you ever seen a robot walk along a waterfront, understand its surroundings, and look out at the sea? 🤖 Thanks to Over the Reality, this is possible. Places like this are mapped every day by our global community of 360 mappers. Using our technology, they are transformed into high-fidelity machine-readable 3D Gaussian Splats. This data can be used to: • navigate robots through the physical world • train AI and Visual Models • power VPS localization • anchor persistent XR content to real-world locations This is just one of more than 250,000 real-world locations already mapped by the Over the Reality community. And unlike traditional datasets, these places can be explored directly through a smartphone, browser, or XR headset, allowing anyone to navigate and experience them in first person. From city streets and waterfronts to landmarks and cultural sites, we're building one of the world's largest machine-readable maps for AI, robotics, and spatial computing.
0:36

Have you ever seen a robot walk along a waterfront, understand its surroundings, and look out at the sea? 🤖 Thanks to Over the Reality, this is possible. Places like this are mapped every day by our global community of 360 mappers. Using our technology, they are transformed into high-fidelity machine-readable 3D Gaussian Splats. This data can be used to: • navigate robots through the physical world • train AI and Visual Models • power VPS localization • anchor persistent XR content to real-world locations This is just one of more than 250,000 real-world locations already mapped by the Over the Reality community. And unlike traditional datasets, these places can be explored directly through a smartphone, browser, or XR headset, allowing anyone to navigate and experience them in first person. From city streets and waterfronts to landmarks and cultural sites, we're building one of the world's largest machine-readable maps for AI, robotics, and spatial computing.

Over the Reality 🌐

32,061 просмотров • 3 месяцев назад

HOW TO TRACK PEOPLE WITH THEIR IMAGES Many photos taken with modern smartphones contain hidden information called GPS metadata. This data is stored in the image file using a format known as EXIF (Exchangeable Image File Format). Along with camera settings like time, date, and device model, EXIF data can also include the exact GPS coordinates of where the picture was taken. When someone takes a photo with location services enabled, the phone uses the Global Positioning System to record the latitude and longitude. These coordinates are automatically embedded into the image file’s metadata. If the image is shared without removing this metadata, anyone with basic tools or online metadata viewers can extract the GPS coordinates from the file. By entering those coordinates into a map service such as Google Maps, they can see the exact location where the photo was taken. Because of this, an image can unintentionally reveal where a person lives, works, or frequently visits. For privacy and safety, many people remove metadata or disable location tagging before sharing photos online.
1:21

HOW TO TRACK PEOPLE WITH THEIR IMAGES Many photos taken with modern smartphones contain hidden information called GPS metadata. This data is stored in the image file using a format known as EXIF (Exchangeable Image File Format). Along with camera settings like time, date, and device model, EXIF data can also include the exact GPS coordinates of where the picture was taken. When someone takes a photo with location services enabled, the phone uses the Global Positioning System to record the latitude and longitude. These coordinates are automatically embedded into the image file’s metadata. If the image is shared without removing this metadata, anyone with basic tools or online metadata viewers can extract the GPS coordinates from the file. By entering those coordinates into a map service such as Google Maps, they can see the exact location where the photo was taken. Because of this, an image can unintentionally reveal where a person lives, works, or frequently visits. For privacy and safety, many people remove metadata or disable location tagging before sharing photos online.

H4RUK1 KIRA🇯🇵

201,347 просмотров • 6 месяцев назад

BREAKING NEWS !! F-35 COMBAT abilities Turned to DUST ! along with all combat aircraft and Bombers heading to IRAN ! NEWS from Douglas Macgregor tonight has just confirmed what has been rumoured with Russian EW TECHNOLOGIES recently tested over Ukraine and Eastern Europe since late 2024 I've wrote about this day several times over last 2 years, that RUSSIA was testing out a GPS denial / spoofing weapon that makes useless all incoming enemy fighter jets and Bombers.... they go BLIND ! Well here it is >>> Iran CUTS Israel's GPS Signal, F-35s FLY BLIND, IDF Loses Air War, U.S PANICS ! Douglas Macgregor This video breaks down a dramatic escalation in modern electronic warfare, where Iran allegedly disrupts Israeli air operations by targeting GPS-dependent systems. We analyze how advanced aircraft like the F-35 could be affected, the mechanics of spoofing vs. jamming, and the broader implications for global military strategy. The discussion explores whether this signals a turning point in aerial dominance and how nations may adapt to contested electromagnetic environments. Covering technical, strategic, and geopolitical angles, this deep dive reveals why control of the spectrum may redefine future conflicts and reshape power balances across the Middle East and beyond. 00:00​ Introduction & scenario overview 08:00​ How GPS warfare works 16:00​ F-35 vulnerabilities explained 24:00​ Iran’s electronic warfare strategy 32:00​ Global military implications PREVIOUS LINKS
0:49

BREAKING NEWS !! F-35 COMBAT abilities Turned to DUST ! along with all combat aircraft and Bombers heading to IRAN ! NEWS from Douglas Macgregor tonight has just confirmed what has been rumoured with Russian EW TECHNOLOGIES recently tested over Ukraine and Eastern Europe since late 2024 I've wrote about this day several times over last 2 years, that RUSSIA was testing out a GPS denial / spoofing weapon that makes useless all incoming enemy fighter jets and Bombers.... they go BLIND ! Well here it is >>> Iran CUTS Israel's GPS Signal, F-35s FLY BLIND, IDF Loses Air War, U.S PANICS ! Douglas Macgregor This video breaks down a dramatic escalation in modern electronic warfare, where Iran allegedly disrupts Israeli air operations by targeting GPS-dependent systems. We analyze how advanced aircraft like the F-35 could be affected, the mechanics of spoofing vs. jamming, and the broader implications for global military strategy. The discussion explores whether this signals a turning point in aerial dominance and how nations may adapt to contested electromagnetic environments. Covering technical, strategic, and geopolitical angles, this deep dive reveals why control of the spectrum may redefine future conflicts and reshape power balances across the Middle East and beyond. 00:00​ Introduction & scenario overview 08:00​ How GPS warfare works 16:00​ F-35 vulnerabilities explained 24:00​ Iran’s electronic warfare strategy 32:00​ Global military implications PREVIOUS LINKS

𝐃𝐚𝐯𝐢𝐝 𝐙 🇷🇺🇮🇪

76,805 просмотров • 5 месяцев назад

BBREAKING: A German robotics startup from Stuttgart just gave robots imagination: Production robotics system where robots evaluate the long-term consequences of their actions before executing them in live industrial environments. Until now, every production robot optimised actions locally; reacting to what it sees right now. The problem? Small errors early in a sequence compound over time. A slightly off pick leads to a jam three steps later. A marginal placement leads to a collision five steps after that. Cortex 2.0 from Sereact introduces decision-grounded world models directly into live operations. The system evaluates alternative action sequences, predicts how risk accumulates, and estimates the likelihood of entering unrecoverable states, before the robot commits. The world model is trained exclusively on real-world execution data. No synthetic simulation. No approximate environment models. Learned from how robots actually fail, recover, and succeed in production. Works across form factors, pick-and-place arms, dual-arm systems, and humanoids. One intelligence layer, any robot body. The Stuttgart-based company Sereact raised a €25M Series A led by Creandum last year, backed by Air Street Capital Capital, Point Nine 🇺🇦 and angels including Nico Rosberg. They already run some of the most productive AI-driven robotic systems in live warehouse environments, with customers including Daimler Truck AG and Bol. ... Stuttgart, not San Francisco. 🇩🇪 Kudos to Ralf Gulde and team! Credit:
0:39

BBREAKING: A German robotics startup from Stuttgart just gave robots imagination: Production robotics system where robots evaluate the long-term consequences of their actions before executing them in live industrial environments. Until now, every production robot optimised actions locally; reacting to what it sees right now. The problem? Small errors early in a sequence compound over time. A slightly off pick leads to a jam three steps later. A marginal placement leads to a collision five steps after that. Cortex 2.0 from Sereact introduces decision-grounded world models directly into live operations. The system evaluates alternative action sequences, predicts how risk accumulates, and estimates the likelihood of entering unrecoverable states, before the robot commits. The world model is trained exclusively on real-world execution data. No synthetic simulation. No approximate environment models. Learned from how robots actually fail, recover, and succeed in production. Works across form factors, pick-and-place arms, dual-arm systems, and humanoids. One intelligence layer, any robot body. The Stuttgart-based company Sereact raised a €25M Series A led by Creandum last year, backed by Air Street Capital Capital, Point Nine 🇺🇦 and angels including Nico Rosberg. They already run some of the most productive AI-driven robotic systems in live warehouse environments, with customers including Daimler Truck AG and Bol. ... Stuttgart, not San Francisco. 🇩🇪 Kudos to Ralf Gulde and team! Credit:

Ilir Aliu

72,419 просмотров • 6 месяцев назад

Mistral AI Releases Robostral Navigate: An 8B Model Enabling Robots to Navigate Complex Environments Hitting 76.6% on R2R-CE With One RGB Camera. No LiDAR. No depth sensor. No multi-camera rig. Here's how it works. 👇 1. Pointing, not metric commands The model predicts the pixel coordinates of the next target in the camera view, plus the arrival orientation. Working in pixel space keeps it robust to camera intrinsics and world scale. When the target leaves the frame, it falls back to local displacements ("2m forward, 1.5m left, turn 25°"). 2. Grounding-first No open-source VLM base. It starts from Mistral's grounding model (pointing, counting, localization). Navigation emerges once the model knows where things are. → ~400,000 trajectories across 6,000 simulated scenes 3. Prefix-caching for training A tree-based attention mask packs a full episode into one sequence — all time steps in a single forward pass. → 22× fewer training tokens; months of training done in days 4. Online RL on top After supervised training, CISPO adds trial-and-error learning to fight distribution shift from behavior cloning. → +3.2% success rate from RL alone 5. The numbers (R2R-CE, Matterport3D) → 76.6% success on validation unseen → +9.7 pts over best single-camera approach → +4.5 pts over best depth/multi-camera system The key takeaway: state-of-the-art continuous VLN without a sensor stack — grounding-init, pixel-space actions, prefix-cached SFT, and online RL, on one RGB camera. Full analysis: Technical details: Mistral AI Mistral AI for Developers
0:22

Mistral AI Releases Robostral Navigate: An 8B Model Enabling Robots to Navigate Complex Environments Hitting 76.6% on R2R-CE With One RGB Camera. No LiDAR. No depth sensor. No multi-camera rig. Here's how it works. 👇 1. Pointing, not metric commands The model predicts the pixel coordinates of the next target in the camera view, plus the arrival orientation. Working in pixel space keeps it robust to camera intrinsics and world scale. When the target leaves the frame, it falls back to local displacements ("2m forward, 1.5m left, turn 25°"). 2. Grounding-first No open-source VLM base. It starts from Mistral's grounding model (pointing, counting, localization). Navigation emerges once the model knows where things are. → ~400,000 trajectories across 6,000 simulated scenes 3. Prefix-caching for training A tree-based attention mask packs a full episode into one sequence — all time steps in a single forward pass. → 22× fewer training tokens; months of training done in days 4. Online RL on top After supervised training, CISPO adds trial-and-error learning to fight distribution shift from behavior cloning. → +3.2% success rate from RL alone 5. The numbers (R2R-CE, Matterport3D) → 76.6% success on validation unseen → +9.7 pts over best single-camera approach → +4.5 pts over best depth/multi-camera system The key takeaway: state-of-the-art continuous VLN without a sensor stack — grounding-init, pixel-space actions, prefix-cached SFT, and online RL, on one RGB camera. Full analysis: Technical details: Mistral AI Mistral AI for Developers

Marktechpost AI

39,955 просмотров • 2 месяцев назад

Hello, Ghana 🇬🇭 Your Excellency, John Dramani Mahama Hello Africa and the world. We 3Farmate are proud to announce the official launch of FAMA—Ghana's FIRST AI-powered autonomous farming robot for large-scale crop production. Founded in 2021 by Clinton 🛸 (CEO) and Koffi-Cobbin (CTO), the company began in a dorm room at Kwame Nkrumah University of Science and Technology, where its first prototype was developed. Since then, the team has engineered FAMA into a full-scale autonomous robot capable of planting seeds, applying fertilizer, weeding, and spraying across real farm environments. FAMA navigates using a vision-based AI system instead of GPS, allowing it to operate reliably in areas where GPS is unavailable or inconsistent. The robot runs on batteries charged by solar panels while in the field and can operate across uneven terrain, loose and muddy soils, and variable weather conditions. A single operator can oversee multiple robots, each covering 27 to 35 acres per day with sub-85mm planting precision. 3Farmate targets large-scale staple crop producers in Ghana, starting with corn and soybeans. We operate a service model, charging farmers per acre and removing the need for upfront equipment investment. Over 70 farmers and several large-scale crop production companies are currently in discussions, with commercial deployments beginning in the 2026 planting season. Approximately $200,000 has been raised to date, including investment from 776 Foundation (Alexis Ohanian, Reddit co-founder) and a grant from Kosmos Innovation Center Ghana. Our team consists of young engineers specializing in robotics, embedded systems, software, and mechanical design. With 8 major iterations, 60+ field test runs, 100+ cumulative acres covered, and thousands of runtime hours in real farm conditions, FAMA is market-ready to become the ultimate farmer-assistant. We built FAMA right here in Ghana, inspired by Dr. Kwame Nkrumah’s belief that “Africa must industrialize to achieve true independence.” We built FAMA as a true testament to every young engineer in Africa that “IT IS POSSIBLE.” FAMA is designed to operate seamlessly on Ghanaian soil and is adaptable to diverse agricultural environments worldwide. Join us as we drive innovation across global agriculture. We call on the government of Ghana, stakeholders, international organizations, and agri-industry leaders to partner with us in transforming agriculture together. This is not history in the making, because history has already been made, and we thank you for being a part of our journey. On this note, we are officially launched! For more information, visit 3FARMATE – The future is here.
1:50

Hello, Ghana 🇬🇭 Your Excellency, John Dramani Mahama Hello Africa and the world. We 3Farmate are proud to announce the official launch of FAMA—Ghana's FIRST AI-powered autonomous farming robot for large-scale crop production. Founded in 2021 by Clinton 🛸 (CEO) and Koffi-Cobbin (CTO), the company began in a dorm room at Kwame Nkrumah University of Science and Technology, where its first prototype was developed. Since then, the team has engineered FAMA into a full-scale autonomous robot capable of planting seeds, applying fertilizer, weeding, and spraying across real farm environments. FAMA navigates using a vision-based AI system instead of GPS, allowing it to operate reliably in areas where GPS is unavailable or inconsistent. The robot runs on batteries charged by solar panels while in the field and can operate across uneven terrain, loose and muddy soils, and variable weather conditions. A single operator can oversee multiple robots, each covering 27 to 35 acres per day with sub-85mm planting precision. 3Farmate targets large-scale staple crop producers in Ghana, starting with corn and soybeans. We operate a service model, charging farmers per acre and removing the need for upfront equipment investment. Over 70 farmers and several large-scale crop production companies are currently in discussions, with commercial deployments beginning in the 2026 planting season. Approximately $200,000 has been raised to date, including investment from 776 Foundation (Alexis Ohanian, Reddit co-founder) and a grant from Kosmos Innovation Center Ghana. Our team consists of young engineers specializing in robotics, embedded systems, software, and mechanical design. With 8 major iterations, 60+ field test runs, 100+ cumulative acres covered, and thousands of runtime hours in real farm conditions, FAMA is market-ready to become the ultimate farmer-assistant. We built FAMA right here in Ghana, inspired by Dr. Kwame Nkrumah’s belief that “Africa must industrialize to achieve true independence.” We built FAMA as a true testament to every young engineer in Africa that “IT IS POSSIBLE.” FAMA is designed to operate seamlessly on Ghanaian soil and is adaptable to diverse agricultural environments worldwide. Join us as we drive innovation across global agriculture. We call on the government of Ghana, stakeholders, international organizations, and agri-industry leaders to partner with us in transforming agriculture together. This is not history in the making, because history has already been made, and we thank you for being a part of our journey. On this note, we are officially launched! For more information, visit 3FARMATE – The future is here.

MARK OFORIQUAYE

150,152 просмотров • 5 месяцев назад

Once we started to work with large global retailers, we needed a better way to scale this process. Ideally, the staff at the store could do this themselves — rather than us flying our team across the world — and then we could lower the cost and timelines. So we built a self-serve version of our survey app, with a tutorial mode designed for beginners. Over time, we collected millions of data points, and so we were able to develop an algorithm which would auto-correct mistakes. In other words, if the surveyor accidentally placed their ground-truth location in the wrong place on the map, we could use our algorithms to detect it, and correct it. So now we have WiFi, and with and our efforts on producing a high quality survey, we have the best WiFi positioning available. With WiFi on its own, it’s achieving 3 meter accuracy. This is a great foundation to build on. WiFi + Motion data To refine this down to 1-meter accuracy, we realised that we could combine WiFi with the same technology behind self-driving cars and robotics: a motion system called SLAM (Simultaneous Localization and Mapping). SLAM uses the accelerometer, gyroscope and camera system to understand precise device motion. Imagine a car driving through a tunnel, using the motion since its last GPS ping to keep location accurate until it comes out the other side. On a phone, this technology is very reliable, and measures device motion with high precision. But SLAM is measuring motion within its own coordinate space, it’s not aligned with the real world. SLAM tracks the user’s relative motion, like “moved forward 2 meters, then turned left”, but does “forward” mean “north”, or some other direction? It’s not calibrated, so it could mean any location, any direction. We can’t rely on the compass to help us out with this, because phone compasses are notoriously incorrect — everyone knows the frustration of being sent the wrong way down a street. So our job was to align this motion data with the triangulation data we were receiving from WiFi. We designed an algorithm that could simulate every possibility, filter the unlikely scenarios, and hone in your location, using WiFi as an anchor. So WiFi gives us the initial blue dot, SLAM gives us motion, and as the user starts walking and we receive more data, our algorithms can refine location accuracy down to a consistent 1-meter accuracy. We’ve tested these algorithms in many locations, on hundreds of hours of ground-truth data:
0:47

Once we started to work with large global retailers, we needed a better way to scale this process. Ideally, the staff at the store could do this themselves — rather than us flying our team across the world — and then we could lower the cost and timelines. So we built a self-serve version of our survey app, with a tutorial mode designed for beginners. Over time, we collected millions of data points, and so we were able to develop an algorithm which would auto-correct mistakes. In other words, if the surveyor accidentally placed their ground-truth location in the wrong place on the map, we could use our algorithms to detect it, and correct it. So now we have WiFi, and with and our efforts on producing a high quality survey, we have the best WiFi positioning available. With WiFi on its own, it’s achieving 3 meter accuracy. This is a great foundation to build on. WiFi + Motion data To refine this down to 1-meter accuracy, we realised that we could combine WiFi with the same technology behind self-driving cars and robotics: a motion system called SLAM (Simultaneous Localization and Mapping). SLAM uses the accelerometer, gyroscope and camera system to understand precise device motion. Imagine a car driving through a tunnel, using the motion since its last GPS ping to keep location accurate until it comes out the other side. On a phone, this technology is very reliable, and measures device motion with high precision. But SLAM is measuring motion within its own coordinate space, it’s not aligned with the real world. SLAM tracks the user’s relative motion, like “moved forward 2 meters, then turned left”, but does “forward” mean “north”, or some other direction? It’s not calibrated, so it could mean any location, any direction. We can’t rely on the compass to help us out with this, because phone compasses are notoriously incorrect — everyone knows the frustration of being sent the wrong way down a street. So our job was to align this motion data with the triangulation data we were receiving from WiFi. We designed an algorithm that could simulate every possibility, filter the unlikely scenarios, and hone in your location, using WiFi as an anchor. So WiFi gives us the initial blue dot, SLAM gives us motion, and as the user starts walking and we receive more data, our algorithms can refine location accuracy down to a consistent 1-meter accuracy. We’ve tested these algorithms in many locations, on hundreds of hours of ground-truth data:

Andrew Hart

91,047 просмотров • 1 год назад

Special thanks to Google DeepMind for inviting me to try out Genie 3. I'm excited to share my thoughts on this early research prototype and also some of my live recordings below: I spent the whole day playing with the system and when it works, it is truly mind blowing🤯. It is the first neural game engine / world model I have tried that generalizes so well and has long term world consistency. Here’s a couple of examples from my live recording and some thoughts on what it means for the future of gaming, robotics, digital experiences and ASI. Where it shines: - Truly general-purpose and quick startup time. Works exceptionally well for gaming environments but also generalizes to other industrial and real-world scenarios. - It learns physics. Although there are systematic failures even for rigid body physics, it was clear to me that it can learn game engine and non-rigid physics without an underlying engine (and in limit learn from game engines via training data). - It works exceptionally well for stylized environments with characters walking around. This will have implications for concept artists, level designers and game devs. - It is way more fun than video models, indicating that there are high retention consumer experiences waiting to be built with this in the future - Photorealistic walk throughs and drone shots work exceptionally well - Global illumination and lighting works surprisingly well - Visual memory is quite powerful and the same objects approximately remain coherent under occlusion and longer time horizons Open Problems: - Physics is still hard and there are obvious failure cases when I tried the classical intuitive physics experiments from psychology (tower of blocks). - Social and multi-agent interactions are tricky to handle. 1vs1 combat games do not work - Long instruction following and simple combinatorial game logic fails (e.g. collect some points / keys etc, go to the door, unlock and so on) - Action space is limited - It is far from being a real game engines and has a long way to go but this is a clear glimpse into the future. The Future: - It is impressive enough for me to have strong conviction that this is going to disrupt the gaming industry. It is super early days and there are a lot of failures but the writing is on the wall. Lots of challenging scientific, engineering and scaling problems to be solved but it is going to happen in the next 5 years. - This is the final piece before we get full AGI and now I think we are well on our way to truly solve it once something like this is scaled up. In many ways it is more ASI than AGI but this is a matter of definitions. The fidelity and generalizability will reach human-level and quickly surpass humans - People are going to combine this with 3D AI and LLMs to build AAA games.
0:57

Special thanks to Google DeepMind for inviting me to try out Genie 3. I'm excited to share my thoughts on this early research prototype and also some of my live recordings below: I spent the whole day playing with the system and when it works, it is truly mind blowing🤯. It is the first neural game engine / world model I have tried that generalizes so well and has long term world consistency. Here’s a couple of examples from my live recording and some thoughts on what it means for the future of gaming, robotics, digital experiences and ASI. Where it shines: - Truly general-purpose and quick startup time. Works exceptionally well for gaming environments but also generalizes to other industrial and real-world scenarios. - It learns physics. Although there are systematic failures even for rigid body physics, it was clear to me that it can learn game engine and non-rigid physics without an underlying engine (and in limit learn from game engines via training data). - It works exceptionally well for stylized environments with characters walking around. This will have implications for concept artists, level designers and game devs. - It is way more fun than video models, indicating that there are high retention consumer experiences waiting to be built with this in the future - Photorealistic walk throughs and drone shots work exceptionally well - Global illumination and lighting works surprisingly well - Visual memory is quite powerful and the same objects approximately remain coherent under occlusion and longer time horizons Open Problems: - Physics is still hard and there are obvious failure cases when I tried the classical intuitive physics experiments from psychology (tower of blocks). - Social and multi-agent interactions are tricky to handle. 1vs1 combat games do not work - Long instruction following and simple combinatorial game logic fails (e.g. collect some points / keys etc, go to the door, unlock and so on) - Action space is limited - It is far from being a real game engines and has a long way to go but this is a clear glimpse into the future. The Future: - It is impressive enough for me to have strong conviction that this is going to disrupt the gaming industry. It is super early days and there are a lot of failures but the writing is on the wall. Lots of challenging scientific, engineering and scaling problems to be solved but it is going to happen in the next 5 years. - This is the final piece before we get full AGI and now I think we are well on our way to truly solve it once something like this is scaled up. In many ways it is more ASI than AGI but this is a matter of definitions. The fidelity and generalizability will reach human-level and quickly surpass humans - People are going to combine this with 3D AI and LLMs to build AAA games.

Tejas Kulkarni

88,083 просмотров • 1 год назад