Andrea Burns

@GoogleDeepMind PhD @BUCompSci she/her/hers

[Andrea Burns @AndreaDBurns]

@kanishkamisra Congrats!!!

[AK @_akhaliq]

congrats to google

also cool to see the worlds LLMs are tested on a gradio app

[Andrea Burns @AndreaDBurns]

I’m at EMNLP and presenting with @roopalgarg at 11am! Come say hi or message if you want to have a coffee chat sometime :)

[Roopal Garg @roopalgarg]

���Join us for our #ImageInWords poster presentation on Nov 12th, 11:00-12:30pm at #EMNLP
. We'll be diving deep into hyper-detailed image descriptions and the impact on #VisionLanguage models. See you there!👋 #NLProc #ComputerVision @emnlpmeeting @AndreaDBurns @GoogleDeepMind 🧵

[Cohere Labs @Cohere_Labs]

Our Geo Regional Asia group is hosting a panel September 4th with @roopalgarg , @bka3 and @AndreaDBurns ! Be sure to add it to your calendar! 🧠
Learn more: https://t.co/MLnCxrUwKz

[Andrea Burns @AndreaDBurns]

@afeyzaakyurek @DerryTWijaya1 We wish you were here!!

[Andrea Burns @AndreaDBurns]

Come by poster 51!!

[Andrea Burns @AndreaDBurns]

I’ll be presenting at the findings session on 8/12 at 17:45. Come find me to chat at any point or explore Bangkok! 👋🏻👋🏻

[Andrea Burns @AndreaDBurns]

A belated announcement that my final work from my PhD was accepted to ACL Findings! I’ll be presenting on Tell Me What’s Next: Textual Foresight for Generic UI Representations

A new approach to learning UI features that has fully open sourced code, pretrained models, and data!!

[Andrea Burns @AndreaDBurns]

To train our model we build a new dataset on top of prior work, OpenApp, which enables several other baselines. Code, models, and data are made available!! Check it out

[Andrea Burns @AndreaDBurns]

@mohitban47 @UNC Congrats Mohit!!

[Andrea Burns @AndreaDBurns]

Congrats to the team for advancing hyper detailed image descriptions!! Grateful to have been a part of this project - check out the released data on GitHub or Huggingface 🥳🥳

[Roopal Garg @roopalgarg]

📢 Excited to unveil our latest research, ImageInWords (IIW)! 🚀We're pushing the boundaries of image descriptions with a new seeded, sequential, human-in-the-loop approach producing SOTA, articulate, hyper-detailed descriptions.

arXiv: https://t.co/LhZKSqtbXo 🧵1/12

[Samarth Mishra @samarthm44]

#NeurIPS2023 bound tomorrow. Looking forward to discovering new papers and having fun conversations. DM if you'd like to chat!

On Thu (5-7 PM CST), I'll help present our poster
"Learning Human Action Recognition Representations Without Real Humans" https://t.co/8br4A9By42

[Andrea Burns @AndreaDBurns]

I’m presenting in 30 mins! See you in the east foyer poster session - let’s chat :)

[Andrea Burns @AndreaDBurns]

Come chat with me about it on the 8th at 4pm. Also happy to meet for coffee or lunch and make more NLP friends!! say hi 👋📷 and we can talk about vision-language tasks, representation learning, or digital domains like webpages and mobile apps

[Andrea Burns @AndreaDBurns]

I'm presenting our WikiWeb2M work on webpage understanding at @emnlpmeeting this Friday - have you ever wanted text, image, and structure relating them all in one place? we provide it for the first time in a unified multimodal webpage sample 🧵

[Andrea Burns @AndreaDBurns]

Prefix Global can prioritize a subset of webpage inputs to specially attend to the rest, while less important image or text has local attention. This is only made possible by having our webpage structure - no more bag of images or words!!