Vishnu Sharma retweetledi
Junior students who have just started doing research?
Check out the (75 and counting) awesome tips!
github.com/jbhuang0604/aw…
English
Vishnu Sharma
149 posts
Vishnu Sharma
@VishnuDSharma
Researcher @BellLabs Past: PhD @umdcs @RAASumd, @comcast @AmexIndia @iitKgp
New Jersey, USA Katılım Mayıs 2012
586 Takip Edilen355 Takipçiler
Vishnu Sharma retweetledi
Can robots get to the point?🎯
We’re sharing AFFORD2ACT—a way for robots to use semantics from a short prompt to keep only the pixels that matter and then act.
This replaces heavy, dense inputs with a small state for control.
1/6
#Robotics #AI #Manipulation #Affordance #Keypoints
English
Easier to search though. Just looks for acronyms with '3R'.
But keep an eye out of gems like VGGT and MapAnything
English
Blink and there’s a new 3D reconstruction paper. Feels like LLM season, just with fewer parameters and more point clouds.
English
Vishnu Sharma retweetledi
Vishnu Sharma retweetledi
From refrigerators to shirts: Inside the strange world of Albert Einstein’s inventions and patents > tinyurl.com/264ax59l
English
I was telling a colleague about Nvidia's Nemotron and he said that it sounds like mecha-clownfish.
Had to ask chatgpt to imagine this
English
Nice! Even the buildings in the background seem well estimated
Nikhil Keetha@Nik__V__
Meet MapAnything – a transformer that directly regresses factored metric 3D scene geometry (from images, calibration, poses, or depth) in an end-to-end way. No pipelines, no extra stages. Just 3D geometry & cameras, straight from any type of input, delivering new state-of-the-art results 🚀 One universal model enables SoTA for: 🔥 Mono Depth Estimation 🔥 Multi-View SfM 🔥 Multi-View Stereo 🔥 Depth Completion 🔥 Registration … and many more possibilities! – plus everything is metric 🎯 We release code for data processing, training, benchmarking & ablations – everything Apache 2.0! Details & Links 👇
English
Vishnu Sharma retweetledi
💡(Ground)BREAKING news: We officially broke ground on the new Nokia #BellLabs HQ at The HELIX NJ!
In attendance were NJ Governor Phil Murphy (@GovMurphy), Mayor James Cahill, our Nishant Batra, @thierry_klein, @PeterVetter6G, @NBDEVCO, & more...
#NokiaBellLabs100 #HELIXNJ
English
Vishnu Sharma retweetledi
Thrilled to present Sketch-to-Skill at #RSS2025 🎉
📍 USC, Bovard Auditorium
🗓️ June 24 | 11:30am talk + 12:30pm poster
Teach robots by drawing ✏️➡️🤖
w/ @peihong_yu, @singhanukriti , @zahir_mahammad & @ptokekar
Paper: lnkd.in/eGmjke3h
website:lnkd.in/eWEMsZ74
AmishaBhaskar@amishabhaskar
🎉 Excited to share that our paper Sketch-to-Skill is accepted to RSS 2025! Congratulations and huge thanks to my amazing collaborators Peihong Yu, @singhanukriti, @zahir_mahammad, and my advisor @PratapTokekar Let me walk you through it. 🧵👇 #RSS2025
English
Vishnu Sharma retweetledi
I’m hiring multiple PhD students and one postdoc at NTU Singapore, starting in Spring or Fall 2026, to push the frontiers of robot planning, learning, and embodied AI.
Details are available here: yoonchangsung.com/opportunity
Thank you for your support in sharing this opportunity!
#PhDOpportunity #Robotics #NTUsg
English
Vishnu Sharma retweetledi
💡 Check out the highlights from our recent Nokia Bell Labs celebration day at our Murray Hill HQ!
Attendees included NJ Governor Phil Murphy, Tech Exec Cathy Hackl, our Justin Hotard, Nishant Batra, Thierry Klein, Peter Vetter & more.
#NokiaBellLabs100 #BellLabs #Science
English
From reactive to proactive: how we used vision-language models to help robots guess what you need next. One of the most fun projects from my PhD.
raaslab.org/blog/IIAA/
English
@jbhuang0604 Or call it BLIP: Blending with Laplacian Image Pyramid
English
Interesting! I didn't realize how important a video title/packaging is until now. 😮
It's the same video, but with a better packaging, it gets much more attention.
(the dashed line is when I experimented with a different title.)
Jia-Bin Huang@jbhuang0604
How a 40-Year-Old Trick Solves Seamless Image Blending Laplacian pyramid blending is a simple yet effective tool for many applications, including object composition, seamless panorama stitching, and exposure fusion. Let’s learn this classic method that still works so well today.
English
A glimpse into some interesting applications of MAE from my work with @singhanukriti and @ptokekar:
vishnudsharma.medium.com/seeing-the-uns…
English
@dhaivat1729 Kokoro is good and available at huggingface. Bark is also nice - it allows injecting extra stuff like laughs with text. You can also try other models from the TTS Arena leaderboard: huggingface.co/spaces/TTS-AGI…
English
What is the best open source text to speech model that could generate natural sounding voices?
English
Vishnu Sharma retweetledi
🚀 Proud to announce two talented graduates, Ruthwik @RuthwikDasyam and Zahir @zahir_mahammad, from our UMD RAAS Lab are now on the market! 🎓🤖
Check out the thread below, for more details on their work! ⬇️
#Robotics #ReinforcementLearning #AI #Graduation #UMD
English
Vishnu Sharma retweetledi
🎂 It's official: Bell Labs is one hundred!
It's hard to imagine what the world would be like without Bell Labs. Join us this year as we celebrate a hundred years of Bell Labs and look forward to the next century of innovation...
Discover more ➡️ nokia.ly/4hHk4lY
English
Probably late to the party but absolutely love the Harry Potter spells on Google Nest. 'Nox' on my way out and 'Lumos' when I come back. Fun and efficient
English