anuragxel Profile Banner
Anurag Ghosh Profile
Anurag Ghosh

@anuragxel

Followers
476
Following
26K
Media
10
Statuses
479

PhDing @CMU_Robotics. CS @iiit_hyderabad. ex-@MSFTResearch. Trying to bring Science Fiction to Real Life.

Pittsburgh, PA
Joined April 2015
Don't wanna be here? Send us removal request.
@anuragxel
Anurag Ghosh
12 days
Very nice and relevant paper for 3D printing and manufacturing!.
@yongyuanxi
Towaki Takikawa / 瀧川永遠希
13 days
You Can't Manufacture a NeRF - an ICML position paper from CMU that discusses what it means to generate a 3D object that can be manufactured. Nothing too controversial and feels a bit more like a survey- but good reference to have for explaining this concept 🙂
Tweet media one
0
0
3
@anuragxel
Anurag Ghosh
1 month
RT @kvuongdev: Will be presenting our poster shortly at 📍#59!. #CVPR2025.
0
10
0
@anuragxel
Anurag Ghosh
1 month
RT @ducha_aiki: Yay! WxBS result in AerialMegaDepth :).#CVPR2025
Tweet media one
Tweet media two
Tweet media three
0
9
0
@anuragxel
Anurag Ghosh
3 months
RT @DomDiTommaso: Pittsburgh right now, apparently:
Tweet media one
0
72
0
@anuragxel
Anurag Ghosh
3 months
Very cool work by @kvuongdev ! 3D prediction for extreme viewpoints is tough. AerialMegaDepth leverages simulated and real data to solve this challenge. A cool result: An aerial image can serve as a "map" for precise localization on the ground even for non-overlapping images.
@kvuongdev
Khiem Vuong
3 months
[1/6] Recent models like DUSt3R generalize well across viewpoints, but performance drops on aerial-ground pairs. At #CVPR2025, we propose AerialMegaDepth (, a hybrid dataset combining mesh renderings with real ground images (MegaDepth) to bridge this gap.
0
0
2
@anuragxel
Anurag Ghosh
3 months
RT @zhenjun_zhao: AerialMegaDepth: Learning Aerial-Ground Reconstruction and View Synthesis. @kvuongdev, @anuragxel, @RamananDeva, Srinivas….
0
13
0
@anuragxel
Anurag Ghosh
4 months
RT @nikhilabm: We taught a robot to cook Michelin-quality dishes. Meet Zippy, the robotic chef:. 🧑‍🍳 Already cooking for Michelin-star che….
0
592
0
@anuragxel
Anurag Ghosh
4 months
RT @LumaLabsAI: Today, we release Inductive Moment Matching (IMM): a new pre-training paradigm breaking the algorithmic ceiling of diffusio….
0
226
0
@anuragxel
Anurag Ghosh
4 months
RT @panagis21: Brought my dad to work today. A retired union painter, all he did was stare at my degrees 😊 . Degrees that wouldn’t be possi….
0
14K
0
@anuragxel
Anurag Ghosh
4 months
RT @Thom_Wolf: I shared a controversial take the other day at an event and I decided to write it down in a longer format: I’m afraid AI won….
0
514
0
@anuragxel
Anurag Ghosh
4 months
RT @gandhikanishk: New Paper!! We try to understand why some LMs self-improve their reasoning while others hit a wall. The key? Cognitive b….
0
184
0
@anuragxel
Anurag Ghosh
5 months
RT @pfau: Maybe my most libertarian view is that you need to let kids play with dangerous stuff if you want to have a society full of great….
0
23
0
@anuragxel
Anurag Ghosh
6 months
RT @POTUS46Archive: One more selfie for the road. We love you, America.
Tweet media one
0
43K
0
@anuragxel
Anurag Ghosh
7 months
RT @lateinteraction: Reminder to self: Much like a good paper always starts with (i) the status of the field at the time of the proposal, (….
0
10
0
@anuragxel
Anurag Ghosh
7 months
Absolutely fantastic work -- an analytical theory for diffusion models!.
@MasonKamb
Mason Kamb @ICML
7 months
Excited to finally share this work w/ @SuryaGanguli. Tl;dr: we find the first closed-form analytical theory that replicates the outputs of the very simplest diffusion models, with median pixel wise r^2 values of 90%+.
Tweet media one
0
0
6
@anuragxel
Anurag Ghosh
8 months
RT @womenpostingws: In 2016, Pharrell Williams visited a NYU music production class to critique songs. After he listened to a song called “….
0
8K
0
@anuragxel
Anurag Ghosh
9 months
RT @DanielleFong: guys, you keep trying to recreate ilya. we can’t. but we can recreate him in the aggregate
Tweet media one
0
31
0
@anuragxel
Anurag Ghosh
9 months
RT @SpaceX: Thousands of distinct vehicle and pad criteria had to be met prior to catching the Super Heavy booster. Thanks to the tireless….
0
10K
0
@anuragxel
Anurag Ghosh
9 months
RT @Abhinav95_: Announcing MatMamba - an elastic Mamba2🐍architecture with🪆Matryoshka-style training and adaptive inference. Train a single….
0
54
0
@anuragxel
Anurag Ghosh
9 months
RT @qunomaan: 1/ 🎉 How can we develop methods to generate synthetic, photorealistic data for training #AI models in robotics? We present Sp….
0
50
0