James Emilian

PhD @CMU_Robotics | DARPA Triage & visual reasoning @AirLabCMU | Prev: @cmu_bme, @fastdotai, mech design @Pentair

[James Emilian @JamesEmilian2]

huh 🤥

[James Emilian @JamesEmilian2]

@JayKarhade over a glass of acai smoothie - 3D/4D is still alive; but I do humanoids to save my mental health 🗣️

[Jia-Bin Huang @jbhuang0604]

POV: attending CVPR

[Dmytro Mishkin 🇺🇦 @ducha_aiki]

Get ready to Image Matching Workshop 2026 today at @CVPR !
Room: 504
13:00 local time
3 star speakers:
⭐ @pesarlin
⭐ @Nik__V__
⭐ @jianyuan_wang
✅10 papers
✅IMC2026 story
#CVPR2026

[F. Güney @ftm_guney]

one-slide rebuttal for explicit models of the world by @ShenlongWang in favor of “vision science” as coined by Bill Freeman earlier in the morning.

also, this is the funniest thing I’ve seen today 😄

[Omar Alama عمر الأعمى @OmarAlama]

Fast & light 2D and 3D zero-shot open-vocabulary semantic segmentation is here 🚀🪶!!

Meet RADSeg:
- 6-30% mIoU improvement while being 3.95x faster and using 2.5x fewer parameters.

- Outperforms combinations of huge vision models (850-1350M) with just 105M !

💡The key is building on the agglomerative model, RADIO, and improving spatial consistency.

[Seungchan Kim @seungchankim25]

We introduce RAVEN, a 3D open-set memory-based behavior tree framework for aerial outdoor semantic navigation. RAVEN not only navigates reliably toward detected targets, but also performs long-range semantic reasoning and LVLM-guided informed search

[Vaishnavh Nagarajan @_vaishnavh]

Wrote my first blog post! I wanted to share a powerful yet under-recognized way to develop emotional maturity as a researcher:

making it a habit to read about the ✨past ✨ and learn from it to make sense of the present

[James Emilian @JamesEmilian2]

reasoning models, GRPO, etc. are still very cool and useful - but throw the a simple train set case from @arcprize at them and it's really really interesting to see how quickly (and confidently) they fail.

[James Emilian @JamesEmilian2]

not-so-hot take - seems to me that LLMs can't truly reason.

[AirLab @AirLabCMU]

🚀 Thrilled to present ViSafe, a vision-only airborne collision avoidance system that achieved drone-to-drone avoidance at 144 km/h.

In an era of congested airspace and growing autonomy, reliable self-separation is paramount 🧵👇

[James Emilian @JamesEmilian2]

From 'The Economics of Scientific Research' - Terrence Kealey (1997).

[James Emilian @JamesEmilian2]

When you mess up your search prompt so bad @Google thinks you’re trying to find a song 😂

[James Emilian @JamesEmilian2]

randomly remembered this - what a fun day! @vincent_hus pls come talk at @cmu_bme :)

[Vincent Hus @vincent_hus]

In a hackathon, every bug fix is like casting another spell to turn chaos into magic....or at least keep it from crashing! Had so much fun!

[James Emilian @JamesEmilian2]

I'd like to imagine this is just GPT having a mini spiral into some inference space that's existential and melancholic; the alternative is scary

[Josie Kins @Josikinz]

I asked chatgpt's new image model to script and generate a series of comics starring itself as the main character. The results genuinely gave me chills.
I'll post them all in a thread below.

[James Emilian @JamesEmilian2]

brb adding ‘sensitive to mirror invariance’ to my resume skills sec :)

[\newcommand{\femb0t}{ @__femb0t]

Sapir Whorf is real ?__?

[arun @xprunie]

@SpaceX @elonmusk

[James Emilian @JamesEmilian2]

@CMU_Robotics @Nishanth_TA @radiology_rsna @MassGeneral @LSUHealthNO There's other projects with lots of cool novel results; which I can talk about there once they're published :)
🚨 sidenote: @Nishanth_TA is on the market as a potential hire for 2025 summer internships; so if anyone reading this is looking for an AI researcher, reach out!

[James Emilian @JamesEmilian2]

'Tis the season of giving, so I'm here on X to highlight my fellow researchers @CMU_Robotics. A 🧵:

[James Emilian @JamesEmilian2]

@CMU_Robotics @Nishanth_TA @radiology_rsna @MassGeneral https://t.co/hziac3Ro9Y

On that theme, I've had the pleasure of collaborating with him on a project for @LSUHealthNO, where we are working on developing new uncertainty quantification paradigms that disentangle aleatoric from epistemic.