The Metaverse Society’s Post

🚀 Excited to announce new SAM 2.1 model checkpoints & the SAM 2 Developer Suite! 🤖 We’re releasing full training/fine tuning code for SAM 2 so you can customize it for your use case with your own data. This was one of the most highly requested features after we launched SAM 1! For use cases that operate on specialized image domains such as medical images or satellite footage, SAM 2 may not work perfectly out of the box -- now you have the tools to improve the model performance by fine tuning on your own data. We're curious to to hear what domain specific models you create! In addition, we know SAM 2 has limitations in certain settings such as missing fine details in fast moving objects (more details on the blog https://lnkd.in/gM5vGA2Z). Now with access to the full training code, we're excited to see the community build on this work and help fix some of these limitations. 💻 For the first time we’re publishing the frontend & backend code for our SAM 2 web demo! For SAM 1 and SAM 2 we built a high quality public demo to allow anyone regardless of ML expertise to try out the model on their own input. This was instrumental in enabling people from diverse backgrounds to apply SAM 2 to their use case. We heard from many people who wanted to build on top of the SAM 2 demo and integrate with other tools and models. We're excited to see what the community builds particularly in the space of creative tools for image and video editing! Please share your creations with us! 💪 We're dropping several new and improved model checkpoints. In SAM 2.1, we trained on longer sequences (16 frames), add additional data augmentation to simulate visually similar & small objects, and improve occlusion handling through some tweaks to positional encoding of spatial/object pointer memory. All checkpoints and code are now available on GitHub: https://lnkd.in/gQJUKvnc

Will 2025 be the year spatial computing takes off? Join our FREE webinar to find out! Spatial computing has been one of the most transformative developments in technology over the past year. But what lies ahead? 🗓 Date: Tuesday, December 17th ⏰ Time: 16:00 GMT Join CCS Insight for an engaging webinar where we'll delve into the key trends, insights and critical factors in the spatial computing market. What to expect: 🔘 Exclusive insights from our latest spotlight report. 🔘 Insights into the current state of the spatial computing market. 🔘 A look at the leading brands and most compelling use cases for VR and AR. 🔘 Predictions for the future trajectory of spatial computing hardware and software. 🔘 Analysis of what's needed for spatial computing to achieve mass market adoption. Led by Leo Gebbie, Principal Analyst and Director, Americas, this session will equip your organisation with the insights you need to navigate the spatial computing revolution: https://lnkd.in/et3VrC_g A replay will be available to all registrants who are unable to attend the session. #SpatialComputing #VR #AR #Metaverse #Innovation #TechTrends #Webinar #CCSInsight

I recently chatted with Dave Carr about his groundbreaking project Neolyfe and his incredible 5-year journey in the decentralized, Web3 space. From his pioneering days at Decentraland to shaping the future of commerce and entertainment through spatial computing, Dave's insights are nothing short of visionary. We touch on RTFKT, Tommy Hilfiger, Dot.Swoosh, IKEA, Roblox, Starbucks, Amazon, Apple to identify all the learnings and how to utilize those for the next phase. 🔍💡 If you're curious about the cutting-edge digital landscapes, you'll want to hear what he has to say about the potential of these evolving technologies. #Innovation #Web3 #SpatialComputing #Neolife #DigitalTransformation

Spatial Learning: The New Frontier in Corporate Education Spatial computing is 2024’s hot tech topic. Alongside emerging technology drivers like genAI, spatial computing solutions—including AR/VR/MR technologies—are changing the workplace today. Vendors and solution providers work closely with enterprise f... https://lnkd.in/eR9mPEEC #AR #VR #Metaverse

Spatial Learning: The New Frontier in Corporate Education Spatial computing is 2024’s hot tech topic. Alongside emerging technology drivers like genAI, spatial computing solutions—including AR/VR/MR technologies—are changing the workplace today. Vendors and solution providers work closely with enterprise f... https://lnkd.in/eR9mPEEC #AR #VR #Metaverse

Alhamdulillah. Another one of the projects from my Master's Thesis at TTU has been published in IEEE Transactions on Human-Machine Systems. We evaluated the efficacy of a physics-based VR system that provides users with realistic cutaneous and kinesthetic haptic feedback. For the evaluation, we used a custom-designed questionnaire, the NASA-TLX, and electromyography activities from biceps, middle, and anterior deltoid muscles to determine user experience, workload, and neuromuscular dynamics, respectively while participants performed sequential pick-and-place lifting tasks. S. T. Mubarrat, A. S. Fernandes, and S. Chowdhury, "A Physics-Based Virtual Reality Haptic System Design and Evaluation by Simulating Human-Robot Collaboration," in IEEE Transactions on Human-Machine Systems https://lnkd.in/esy6YpeN Public access: https://lnkd.in/eCm2jy-G Related article: S. T. Mubarrat, S. Chowdhury and A. S. Fernandes, "Evaluating Visual-Spatiotemporal Co-registration of a Physics-based Virtual Reality Haptic Interface," in IEEE Access https://lnkd.in/g-iyCeFv

Spatial Learning: The New Frontier in Corporate Education Spatial computing is 2024’s hot tech topic. Alongside emerging technology drivers like genAI, spatial computing solutions—including AR/VR/MR technologies—are changing the workplace today. Vendors and solution providers work closely with enterprise f... https://lnkd.in/eR9mPEEC #AR #VR #Metaverse

The Spatial Computing: Revolutionizing Industries with AI, AR, and VR Integration - Key takeaways #spatialcomputing #AI #VR #AR #mixedreality #immersivetechnology #humancomputerinteraction #espincorp https://lnkd.in/gAcuaMFi

Imagine walking freely in endless virtual world, without worrying about physical space. 🌐 Charlotte Croucher’s PhD research is tackling the “locomotion problem” in VR with redirected walking (RDW) controllers. These controllers guide users safely in physical spaces while navigating expansive virtual environments. Charlotte developed the LoCoMoTe framework and dashboard, a tool that bridges machine learning simulations with live-user experiments, paving the way for open science and collaboration. Good luck, Charlotte, on defending your thesis today! Read more about her research here: https://lnkd.in/edw3CVni 

Discover how AR/VR is transforming middle school science! From dissecting virtual frogs to exploring the solar system, see how these immersive technologies are sparking curiosity and enhancing learning. Dive into our latest blog now: https://hubs.li/Q02BZx0R0