Ananth Kalyanasundaram

CS graduate student @ TUM
Computer Vision, Deep Learning

[Ananth Kalyanasundaram @ananthk2603]

switched to Codex from Claude code. Love that there isn't a silent downgradation/routing of models when I use 5.6-Sol and has generous limits.

[SpAItial AI @SpAItial_AI]

We❤️Gaming 💕

AI-generated first-person shooter made easy: input an image -> Echo-2 generates world -> let's play!

We're already having fun playing but it'll be open soon :)

[Ananth Kalyanasundaram @ananthk2603]

This is so badass! 🔥 @MattNiessner

[Matthias Niessner @MattNiessner]

📢𝐁𝐈𝐆 𝐍𝐄𝐖𝐒: 𝐋𝐚𝐮𝐧𝐜𝐡𝐢𝐧𝐠 𝐄𝐜𝐡𝐨-𝟐 𝐓𝐨𝐝𝐚𝐲📢

My obsession with virtual environments started with childhood video games. But after years of research in 3D reconstruction and neural rendering, the bottleneck became obvious: we don't just need to generate better pixels. We needed a foundational model that natively understands space and the underlying physics.

That spatial grounding is exactly what you are seeing in the thread below.

Echo-2 enables a two-way flow of knowledge between reality and simulation. It is the bridge between capturing the physical world and building the high-fidelity simulations required to train tomorrow's robots.

"What I cannot create, I do not understand." — Richard Feynman.

[SpAItial AI @SpAItial_AI]

🚀Echo-2 is here - our new world model!

These aren’t videos. These are 𝟑𝐃 𝐬𝐜𝐞𝐧𝐞𝐬. Generated from a single image.

- Stunning visual quality.
- Real-time rendering.
- Interactive camera control.
- Physically grounded.

🧵More details👇

[Ananth Kalyanasundaram @ananthk2603]

Pretty cool that Grok 4.1 can create OpenGL simulations of blackholes now. ❤️‍🔥 @elonmusk

[Ananth Kalyanasundaram @ananthk2603]

@LinusTech Check out my first build. RTX 3090 + i9-14900K and a LG Ultragear QHD display.

[Ananth Kalyanasundaram @ananthk2603]

360mm AIO*

[Ananth Kalyanasundaram @ananthk2603]

Have always dreamed of buying a high end setup as a big fan of @LinusTech @linusgsebastian .
Was able to realize that dream yesterday. Wow @nvidia RTX 3090 is a big chunky one! Love every minute of using it.

P.S: How you do cable management with a BFGPU and 360mm is beyond me.

[Matthias Niessner @MattNiessner]

(1/3)
Can we turn text-to-image models into photorealistic 3D generators?

ViewDiff (#CVPR2024) produces realistic, multi-view consistent images of real-world 3D objects in authentic surroundings.

Website https://t.co/ukm8rIJXU5
Video https://t.co/7NTrMxveJg

How does it work?

[Ananth Kalyanasundaram @ananthk2603]

@overleaf Hey, looks like the server is down again. :/

[Matthias Niessner @MattNiessner]

Our TU Munich campus in Garching just turned into a winter wonderland 🏂🎿🏔️☃️

[Matthias Niessner @MattNiessner]

Great research is often a *very* simple idea.

But don't be fooled: finding out which idea actually works, can be extremely difficult. Obviously, once it's there and everyone uses it, everything is easy :)

[Matthias Niessner @MattNiessner]

Happy to share the #SIGGRAPH'2023 paper compilation from our lab & collaborators:
https://t.co/1O101FsavB

We present exciting ideas on generative 3D models (diffusion & GANs) for 3D shapes, faces, and humans :)

Looking forward to a great conference in #LosAngeles next week!!!

[Matthias Niessner @MattNiessner]

Can we apply diffusion directly on MLP weights? Yes!!!

We show generalization to new 3D shapes and 4D mesh animations: "HyperDiffusion: Generating Implicit Neural Fields with Weight-Space Diffusion" #ICCV2023!

Project: https://t.co/SqECW9MOmY
Video: https://t.co/xkIdJcg5M4

[Matthias Niessner @MattNiessner]

(1/4)
Excited to share our #ICCV2023 paper Text2Room!

We generate scene-scale textured 3D meshes from a given text prompt leveraging 2D text-to-image models such as StableDiffusion.

Project: https://t.co/X8fLEIDnRu
Code: https://t.co/QMpoyGQYRS
Video: https://t.co/l6UNFWQ46S

[Yann LeCun @ylecun]

The US is notorious for it's mass shootings, but its immigration policies (or lack thereof) set a new standard for the art of shooting yourself in the foot with a bazooka.

Advice to graduate students from countries the US doesn't like: just go to Europe.

[Matthias Niessner @MattNiessner]

We released our ActorsHQ dataset from our #SIGGRAPH work: "HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion".

ActorsHQ provides high-fidelity multi-view captures at 12MP resolution from 160 cameras with per-frame mesh reconstructions.

https://t.co/WwoOBwKUYj

[Matthias Niessner @MattNiessner]

(1/2)
Happy to announce that Text2Tex has been accepted at #ICCV2023 🎉

Taking a mesh and a text prompt as input, Text2Tex generates high quality textures - it's fully automated and easy to scale to many models!

Project: https://t.co/6Hp9zGsaPH
Video: https://t.co/fu0awrihBE

[Ananth Kalyanasundaram @ananthk2603]

@yinyu_nie 🥳

[Matthias Niessner @MattNiessner]

Our "Introduction to Deep Learning" (I2DL) 2023 lecture is now on YouTube!

We cover DL fundamentals - backprop, SGD, losses, CNNs, etc. - as well as advanced concepts such as RNNs/LSTMs transformers, generative models (GANs, diffusion)!

Check it out: https://t.co/SceSnA75T9