Shagunn (@shagunuppls) / X

Shagunn

113 posts

Shagunn

@shagunuppls

Brainstorming with bots 🤖🧠 @SkildAI || MS Robotics @CarnegieMellon @CMU_Robotics || Previously: @Preimage_AI | CLVR Lab @USC | BRAIN Lab @SUTDsg | @iiitdcai

shagunuppal.github.io

Joined May 2013

Pinned
Shagunn
@shagunuppls
May 29, 2024
Want a robot to navigate a cluttered room and fetch you something? Presenting SPIN: seamlessly moves past obstacles using active vision & whole-body coordination. SPIN learns an end-to-end policy in simulation, jointly optimizing perception & action spin-robot.github.io 1/n
00:00
77K
Shagunn
@shagunuppls
Jul 9, 2024
Extremely stoked and fortunate to be a part of this journey with such incredible mentors @pathak2206, @gupta_abhinav_ and a stellar team @SkildAI ! 🚀Could not be more excited about what lies ahead while we unfold our vision of building general purpose embodied AI.
Deepak Pathak
@pathak2206
Jul 9, 2024
We’re striving to build the most scalable, robust, and generalizable robotics foundation model. Thank you to our team for all their hard work so far & our investors for believing in our mission. Follow along with us @SkildAI for the latest. Blog: skild.ai/blogs/announci…
8.1K
Shagunn
@shagunuppls
Jul 3, 2020
Our paper titled 'Disentangling Multiple Features in Video Sequences using Gaussian Processes in Variational Autoencoders' with @sarthak__bhagat, Vivian Yin, Nengli Lim got accepted at #ECCV2020. Preprint available at arxiv.org/abs/2001.02408.
Shagunn
@shagunuppls
Jan 26, 2021
It being my first ever publication as an undergrad, makes it even better! :D
Sarthak Bhagat
@sarthak__bhagat
Jan 26, 2021
Today one of my papers received a citation from @AndrewYNg's paper. For someone whose very first interaction with ML was through his @coursera lectures, this feels huge. @shagunuppls
Shagunn
@shagunuppls
Aug 24, 2020
We will be presenting our paper "Disentangling Multiple Features in Video Sequences using Gaussian Processes in Variational Autoencoders" (Poster #2183) at #ECCV2020 . Join zoom Q&A session on August 25 from 00:00-02:00 and 06:00-08:00 (UTC+1) to discuss.
Brandlive
From brandlive.com
Shagunn
@shagunuppls
Aug 9, 2024
Imagine a robot that can assist you in the kitchen just by observing your actions, no instructions needed! 🤖🏠 Check out this amazing work led by @sarthak__bhagat!
Sarthak Bhagat
@sarthak__bhagat
Aug 9, 2024
Want a robot to assist you in the kitchen **without any instructions** simply by watching you?🤖🏠 🚀 Presenting our recent paper on action anticipation from short video context for human-robot collaboration, accepted at Robotics and Automation Letters (RA-L).
00:00
3.2K
Shagunn
@shagunuppls
Jun 18, 2024
Excited to present our work SPIN at the Embodied AI Workshop from 1-2 PM on Tuesday as well as the @CVPR conference from 5:15-6:45 PM on Thursday. Feel free to join if you are interested in sim2real and RL for mobile manipulation.
Deepak Pathak
@pathak2206
Jun 17, 2024
Reactive vs Planning is a big debate. As a proponent of bottom-up robotics, I am always curious as to how far we can push the reactive paradigm. Presenting SPIN this week at #CVPR2024: end-to-end policy for perception, navigation & manipulation without any mapping or planning.
1.7K
Shagunn
@shagunuppls
Aug 28, 2020
Our work (with @sarthak__bhagat,@vishaal_urao,Dr. Saket Anand) on self-supervised visual disentanglement was accepted at #ECCV2020 Workshop - Perception Through Structured Generative Models (PTSGM). Join us at us04web.zoom.us/j/79248924928?… for Q&A at 14:00 and 22:00 (UTC+1). @eccvconf
Shagunn
@shagunuppls
Mar 22, 2021
Excited to be a part of it! :)
WiMLDS Delhi
@WiMLDS_Delhi
Mar 22, 2021
Announcing the mentors! We are thrilled to have @shagunuppls on-board!
Shagunn
@shagunuppls
May 29, 2024
Replying to @shagunuppls
We must perceive in order to move but we must also move in order to perceive! Our robot adapts in real-time, navigating even dynamic obstacles through emergent behavior. Be it a crowded workspace or just a dirty room! 😛 Paper: arxiv.org/abs/2405.07991 6/6
00:00
11K
Shagunn
@shagunuppls
Dec 21, 2023
Combining sim2real with Internet data for large-scale generalization, a low-cost dextrous robot hand adeptly grasps diverse tools such as hammers, drills, and screwdriver to everyday objects like mug, pan and much more!
Ananye Agarwal
@anag004
Dec 21, 2023
Robotic intelligence requires dexterous tool use, but generalizing across tools is hard. Our CoRL23 paper combines semantics (affordances) with low-level control (sim2real) to show functional grasping that generalizes to hammers, drills and more! dexfunc.github.io 1/n
00:00
1.9K
Shagunn
@shagunuppls
May 29, 2024
Replying to @shagunuppls
w/ @anag004, @Haoyu_Xiong_, @kenny__shaw, @pathak2206 No map, no problem!📍🚫 Just as toddlers or animals navigate new environments with ease almost like muscle memory, SPIN adapts & reacts on the move. With only ego-vision, it shows agile local mobility & obstacle avoidance 2/
00:00
2.6K
Shagunn
@shagunuppls
May 29, 2024
Replying to @shagunuppls
SPIN exhibits emergent behavior like spatial awareness and whole-body coordination. When a floating obstacle suddenly appears, it utilizes its arm to navigate through instead of rerouting its base, demonstrating faster and more efficient navigation strategies. 5/
00:00
1.1K
Shagunn
@shagunuppls
May 29, 2024
Replying to @shagunuppls
In partially observable environments, SPIN showcases selective perception with an actuated camera. It consciously decides what to perceive and leverages joint action-perception optimization to display hand-eye coordination 🙌👁️, all with a single end-to-end policy. 3/
1.4K