Rohan Chacko

3D @NianticSpatial. let it rip

[Niantic Spatial 🌎 @NianticSpatial]

Niantic Spatial is working with @HMCI_AI and the @CityofRCordova to build the foundation for the city’s first digital twin for physical AI.

Roughly four square kilometers of the Innovation Region is already reconstructed and running in simulation.

Read more: https://t.co/NlICQakw4G

#NianticSpatial #PhysicalAI #DigitalTwins #SpatialAI #3DReconstruction #Robotics

[Rohan Chacko @rchacko23]

@pablovelagomez1 @rerundotio Finally! Been waiting so long for this!!

[Pope Leo XIV @Pontifex]

We cannot consider #AI to be morally neutral. In reality, every technical tool embodies choices and priorities through what it measures, ignores, and optimizes, and how it classifies people and situations. Ethical discernment cannot be limited to asking whether we are using a system for good or bad purposes. It must also examine how that system is designed and what vision of the human person and society is embedded in the data and models that guide it. #MagnificaHumanitas

[LeRobot @LeRobotHF]

Is it really worth it joining a robotics hackathon?

One year ago, thousands of people around the world spent a single weekend building AI robots using the open-source LeRobot stack. Many had never built a robot before. We sat down with the winners to ask what they made and where they are now.

For many of them, that weekend changed everything, turning the intimidating field of AI robotics into an accessible hobby or even a full-time career.

Here are a few key takeaways from their year-long journey:
- You don't need a PhD to start. For many participants, this was their very first robot
- One weekend can launch a career. Winning secured internships and full-time jobs
- Community lowers the barrier to entry
- Learning by doing is exponentially faster. Robotics still requires getting your hands dirty to truly progress

Watch the full video to see their projects and hear their advice for taking the first step.

@ClementDelangue, @Thom_Wolf

[Rohan Chacko @rchacko23]

@Azadux 👀

[Rohan Chacko @rchacko23]

@jxnlco No

[asim ᯅ @asimahmed]

for physical AI to work in the real world, it needs high-fidelity 3D models of the real world.

not just for visualization — for localization, spatial reasoning, and real-to-sim workflows that close the gap between training and deployment.

at @nianticspatial, we're turning @spexigeospatial drone imagery into exactly that — city-scale Gaussian splats grounded in the environments that matter most.

[Rohan Chacko @rchacko23]

@DuckbillStudio Looks beautiful!

[Rohan Chacko @rchacko23]

@sparkjsdev Great work! This is pretty cool!

[asim ᯅ @asimahmed]

another high-fidelity gaussian splat i took recently. 2 mins scan with 360° camera.

in the splat we simulate how VPS localizes robots to the real world, mirroring the 360° video of the robot navigating in the physical location.

powered by @nianticspatial.

[Niantic Spatial 🌎 @NianticSpatial]

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: https://t.co/9oMKFBCnLA

#NianticSpatial #Scaniverse #VPS #GeospatialAI #AI

[Casey Neistat @Casey]

did a project w meta quest 3 where i scanned my studio. its pretty rad, you can walk around a look at stuff and get close. i spefically did not clean the place before we scanned it. works on your phone and on the headset - https://t.co/1sVyNrq5MB

[ChessBase India @ChessbaseIndia]

Divya Deshmukh beats Koneru Humpy in tiebreaks to become the FIDE Women's World Cup Champion 2025 - and become India's 88th Grandmaster!

In the all-Indian Finals which went to tiebreaks, Divya defeated Humpy 1.5-0.5. The first Rapid game ended in a draw, and the next one Divya scored an amazing win with the Black pieces to win the match!

Divya qualified to the Candidates Tournament by entering the finals. Now she becomes the first Indian to win the FIDE Women's World Cup, and directly gets her Grandmaster title as a result!

A Women's World Cup Champion, and a new Grandmaster on the same day. What a moment for Indian Chess! Huge congratulations to Divya , her team and her family for this amazing achievement.

Graphic: Anmol Bhargav
#Chess #ChessBaseIndia #divyadeshmukh #fideworldcup

[ElevenLabs @ElevenLabs]

Introducing Eleven v3 (alpha) - the most expressive Text to Speech model ever.

Supporting 70+ languages, multi-speaker dialogue, and audio tags such as [excited], [sighs], [laughing], and [whispers].

Now in public alpha and 80% off in June.

[Techstars @Techstars]

Networking matters! InstaLabel (Techstars 2024) landed investments from 3 unicorn founders through Mentor Madness. 🦄

Watch the full interview:
🔗 https://t.co/xxNp3cB3bu

[Rohan Chacko @rchacko23]

@jianhao75895505 Great work!

[Alexandre Morgand @Almorgand]

How can we leverage 3D Gaussian Splatting for 3D object extraction and 2D segmentation?

"Lifting by Gaussians: A Simple, Fast and Flexible Method for 3D Instance Segmentation" from @magicleap [1/4]

[Rohan Chacko @rchacko23]

@drsrinathsridha @RahulSajnani2 @wacv_official @kapilkatyal_dc @AmazonScience Congrats Rahul! 🚀

[Rohan Chacko @rchacko23]

Our #WACV2025 work on training-free 3DGS instance segmentation

[MrNeRF @janusch_patas]

Lifting by Gaussians: A Simple, Fast and Flexible Method for 3D Instance Segmentation

Contributions:
• A training-free 3D instance segmentation approach that utilizes a novel 2D-to-3D lifting strategy. It assigns Gaussian semantics and employs an incremental merging procedure based on geometric and semantic overlap criteria. This approach enables fast 3D segmentation of any existing 3DGS field without needing costly optimization of the 3D semantic field.

• A 3D segmentation refinement step that allows extraction of visually appealing 3D assets from existing 3DGS fields.

• A novel 3D semantic segmentation dataset and a quantitative evaluation protocol for assessing the fidelity of individually extracted 3D assets.

[Rohan Chacko @rchacko23]

@jon_barron https://t.co/ufzcshc2UY kind of