Meera Hahn

Research Scientist @GoogleAI
PhD in Computer Science @GeorgiaTech
Undergrad @EmoryUniversity

[Zi Wang, Ph.D. @ziwphd]

Check out Proactive Co-Creator on
@GoogleAIStudio , a human-AI belief alignment demo I vibe coded: https://t.co/BEcchP0K89

🧠 See & edit the AI's uncertainty via belief graph. It asks clarifying questions before creating!

📷 Try Image ➔ Story ➔ Video. You can even remix it!

[Meera Hahn @MeeraHahn]

A great example of how creatives and artists can use ai to bring to life previously impossible ideas!

[Google DeepMind @GoogleDeepMind]

Our short film Dear Upstairs Neighbors is previewing at @sundancefest. ���

It’s a story about noisy neighbors, but behind the scenes, it’s about solving a huge challenge in generative AI: control.

Developed by Pixar alumni, an Academy Award winner, researchers, and engineers, here’s how it came together. 🎨

[Meera Hahn @MeeraHahn]

Exciting new work from @sihyun_yu and our team at Google Deep Mind!

Memory-Augmented Latent Transformers (MALT) Diffusion, a new diffusion model specialized for long video generation!
https://t.co/gcDZr5mVbf

[Nithish Kannen @NithishKannen]

Check out our tech report on proactive T2I agents that ask clarification questions to reduce uncertainty!

With this agent, we obtain 2 times higher VQAScore in just 5 turns!🤯 We've open-sourced our agent code powered by #Gemini @GoogleDeepMind! 🚀

Code:https://t.co/67dTGDQpZI

[Nataniel Ruiz @natanielruizg]

legit treflip by Veo 2 with just one error (wheel inversion in the middle). legs move realistically for a treflip 🤯. skateboarding videos are notoriously hard to generate

[Zi Wang, Ph.D. @ziwphd]

Many thanks to @MeeraHahn , Wenjun Zeng, Nithish Kannen, Rich Galt, Kartikeya Badola, @_beenkim for making this happen! Stay tuned for code release!

[Zi Wang, Ph.D. @ziwphd]

Tired of endless prompt tweaking? We've released a tech report on proactive text-to-image agents powered by #Gemini @GoogleDeepMind! Our agents ask clarifying questions and use belief graphs to understand what you really want.

https://t.co/jRwjxtqALx
https://t.co/XFVvva7r96

[Agrim Gupta @agrimgupta92]

6/ Finally, our model can be used to generate videos with consistent 3D camera motion.

[Agrim Gupta @agrimgupta92]

2/ website: https://t.co/atH5wzRudu

Our approach has two key design decisions. First, we use a causal encoder to compress images and videos in a shared latent space.

[Agrim Gupta @agrimgupta92]

We introduce W.A.L.T, a diffusion model for photorealistic video generation. Our model is a transformer trained on image and video generation in a shared latent space. 🧵👇

[Philipp Tsipman @ptsi]

If you’re having trouble keeping up with Video AI😅, there have been 5 state-of-the-art generative video models released *in last 7 days*: 🤯😎🧵

[Meera Hahn @MeeraHahn]

Have you ever wondered about emergent intelligence in robotic agents?

This work shows interesting emergent intelligence and behaviors in blind navigation agents! Blind agents learn maps as they navigate. This allows them to navigate as successfully as an agent with vision

[erikwijmans @erikwijmans]

How do 'map-less' agents navigate? They learn to build implicit maps of their environment in their hidden state!

We study 'blind' AI navigation agents and find the following 🧵

[Max Xu @maxxu05]

How can we fill in missing pulsative sensor data? Prior state-of-the-art fails in our novel setting, despite its well-defined temporal structure.

Checkout our #NeurIPS2022 paper, PulseImpute, @ 4 pm CST!
arxiv: https://t.co/Hbv2x7ZvkP
github: https://t.co/bTManuLyEH

[michelle huang @michellehuang42]

i trained an ai chatbot on my childhood journal entries - so that i could engage in real-time dialogue with my "inner child"

some reflections below:

[Stefan Stojanov @sstj389]

Dense self-supervised learning from multiple 3D viewpoints → dense feature representations that generalize both to novel object instances and to novel categories of instances.

Checkout our #NeurIPS2022 paper!
arxiv: https://t.co/rxzdrScII1
github: https://t.co/5n98W7Wykt

[Meera Hahn @MeeraHahn]

We model indoor environments using FPV panoramic navigation graphs and introduce a visiolinguistic transformer model, LED-Bert, which scores the alignment between navigation graph nodes and dialogs and achieves SOTA performance on the LED task!

[Meera Hahn @MeeraHahn]

✨Transformer-based Localization from Embodied Dialog with Large-scale Pre-training✨ has been accepted as an oral at @aaclmeeting!

https://t.co/L7SSGfqk0w
w/ @RehgJim

[Nataniel Ruiz @natanielruizg]

Today, along with my collaborators at @GoogleAI, we announce DreamBooth! It allows a user to generate a subject of choice (pet, object, etc.) in myriad contexts and with text-guided semantic variations! The options are endless. (Thread 👇)
webpage: https://t.co/EDpIyalqiK
1/N

[Stefan Stojanov @sstj389]

A new year, a new shameless twitter plug: Check out our Toys4K 3D object dataset

4K instances, 105 categories, 15+ instances per category

https://t.co/K85J007iru

[AK @_akhaliq]

No RL, No Simulation: Learning to Navigate without
Navigating
abs: https://t.co/xWFQwed1ES