Chandradeep Pokhariya

PhD Student, Inria Grenoble

[Srinath Sridhar @drsrinathsridha]

My PhD student Rao Fu (@RaoFu79761158) is on the academic job market. She is an expert on creating AI that generates/understands dynamic 3D worlds from language. She is a top student (MIT EECS Rising Star), and will be an amazing colleague!

More here: https://t.co/N96YTGR6RW

[Arthur Chen @ArthurKefan]

Excited to share our latest work, "FoundHand: Large-Scale Domain-Specific Learning for Controllable Hand Image Generation," selected as ⭐Highlight⭐ at #CVPR2025!
🔗 Website https://t.co/YtmwGbYYFq
Thanks to all my amazing collaborators!

[Brown CS @BrownCSDept]

@BrownCSDept PhD student @RahulSajnani2 and co-authors have received a @wacv_official Best Student Paper Award for a novel approach to image editing by unifying common 2D and 3D object editing capabilities into a single method. Learn more: https://t.co/cTAEW5Y7bL

[Srinath Sridhar @drsrinathsridha]

This past winter, I taught a 5-part course on Radiance Fields at @iiscbangalore. We looked at Radiance Fields with a historic lens: 530 million years ago to 2025, and covered both NeRF and 3DGS. Thanks to @KotakIIScAIML and @val_iisc for hosting me!

https://t.co/A5NgQ4wuCI

[Chandradeep Pokhariya @coreqode]

@aashishrai3799 @drsrinathsridha Congrats!! Nice work.

[Srinath Sridhar @drsrinathsridha]

At @wacv_official, we will present EgoSonics, a method to generate audio tracks for silent egocentric videos. A thread 🧵

[Chandradeep Pokhariya @coreqode]

@NateMorrical Oh, that's awesome. I think you might already be aware but there is this work on non-linear sphere tracing to deform SDFs (https://t.co/iuRMOghe1L)

[P J Narayanan @PJNarayanan]

https://t.co/C1EjbrVLiP

@nvidia and Jensen Huang @nvidia_ceo is in the news. They can exceed expectations and attract market's indifference (or wrath).

Here is my take on how Nvidia reached where they did, with some well-deserved appreciation.

[Chandradeep Pokhariya @coreqode]

@Luckyballa Is this constraint-based stable Neo-Hookean (https://t.co/YPA51S55Xm) ?

[P J Narayanan @PJNarayanan]

Our #EGSR2024 paper, with my student @ishaan_n_shah, Adrien Gruson @BeltegeuseCG, and Luis Gamboa, sped up glint rendering using neural histograms. We handled spatial varying surface roughness for the first time. Do check it out!

Adrien presenting it: https://t.co/TFiDCIJrLh

[Chandradeep Pokhariya @coreqode]

@_TheodoreKim @YaleEngineering Congratulations :)

[Chandradeep Pokhariya @coreqode]

@ninnnnnglala excited !!!

[Kiana Ehsani @ehsanik]

[Probably My Most Controversial Twitter Thread to Date] I've been debating whether to write this tweet or not. But I know I'll regret it if I don't say anything. Some of this year's review process decisions were downright bizarre! 🤔 Even in the AC workshop, it was clear some groundbreaking papers got snubbed or accepted without proper recognition (DUSt3R, among others). What's going on here? 🚨📉#CVPR2024

[Makarand Tapaswi @MakarandTapaswi]

📢Happy to announce two #CVPR2024 papers from our Katha AI group @iiit_hyderabad! 🎉🎞️🔥

1. On 📺TV episode story summarization using recaps with @rodosingh23, @sridhruv.
2. On identity-aware video captioning with @HRaajesh, @dnaveenr, @zeeshank95.

Paper 🧵and arXiv soon. 1/4

[Venkatesh Babu @rvbabuiisc]

Presenting 'Balancing Act: Distribution-Guided Debiasing in Diffusion Models' poster in 4E-182 at @CVPR now. Please visit our poster to know more!

[Srinath Sridhar @drsrinathsridha]

We will present MANUS at the poster session today, 10:30 a.m. PDT.

Arch 4A-E. Paper #196.

Please come to talk to @coreqode, @_angelaxing , and me.

[Srinath Sridhar @drsrinathsridha]

Our group will be at CVPR represented by @coreqode, Cheng-You, Peisen, @_angelaxing, and myself. We are presenting two main conference papers, two workshop papers, and three invited talks. Details below and in thread. Looking forward to meeting everyone in Seattle! 🧵

[Srinath Sridhar @drsrinathsridha]

Grasps are one of the primary ways in which we interact with and shape our environments. How can we faithfully capture human grasps with details such as hand/object shape and contact points? At @CVPR, we present MANUS, a method to accurately reconstruct grasps and contacts. 🧵