Pedro F. Proença

Prof. @NOVAFCT / ex-Eng. @NASAJPL / dev defense startups / Creator of 🌱powered

[Pedro F. Proença @flying_pedro]

For the past year, @seomisw and I have been building, in our spare time, an image dataset for litter detection: https://t.co/QbiwNkMulf Our aims are to crowdsource more annotated images and support work on #computervision #deeplearning #AIforGood #plasticpollution #Environment

[Pedro F. Proença @flying_pedro]

For the average european, GDPR may look benign. The truth is it has been a heavy burden on small companies and a source of many contradictions. Can you use car sentry mode?

[DHH @dhh]

The GDPR has been a catastrophe for Europe. The billions of euros misspent on stupid compliance games to with loser prizes like this is tragic. A bold EU would recognize the lateness of the hour and flush it whole.

[Pedro F. Proença @flying_pedro]

Big leap. 3d modelling has been a huge bottleneck.

[Sharif Shameem @sharifshameem]

GPT-6 Astra recreated the Palace of Fine arts in Blender.

This is favorite building in San Francisco because it was built for the World's Fair in 1914 where the steam locomotive and telephone were showed off. It was a time when technology gave us all a deep sense of optimism for the future.

I feel like some of that sense has been lost since, but I hope models like astra can help restore it, and push us towards a more hopeful future.

[Pedro F. Proença @flying_pedro]

First smoke test for our unitree G1 stack

[Pedro F. Proença @flying_pedro]

it’s a really long tail

[Pedro F. Proença @flying_pedro]

🪦

[Tianxing Chen @MarioChan2002]

We evaluated 30+ frontier embodied AI models.
The result is clear: current generalist robot policies are still far from robust real-world manipulation.
This is why we built RoboDojo.

[Pedro F. Proença @flying_pedro]

I saw this before the heat and knew this was gonna trigger a lot of old schoolers - rightfuly. Didn’t take it seriously, but what worries me is that I see this idea spreading and students skipping the classical CV: SE(3), camera models, Kinematics, inertial nav

[Jon Barron @jon_barron]

I'll be giving a "exactly how bitter lesson'ed is all of 3D computer vision?" talk at the Bitter Lessons CVPR workshop tomorrow at 10:30am, Room 3A-3D. Here's a slide with the overall thesis.

[Pedro F. Proença @flying_pedro]

I am excited to join @FCTNOVA Computer science Department as an assistant professor working on spatial AI and autonomous systems. 🚀

[Pedro F. Proença @flying_pedro]

Hires images of Altadena post-fire: https://t.co/pBoKZjy9nN

Can’t believe it’s all gone. So many good memories.

[Pedro F. Proença @flying_pedro]

Examples of issues: Roundabouts, cars parked on the sidewalk, narrow roads, poor map directions, no supercharger maps, no turn signal levers

[Pedro F. Proença @flying_pedro]

Tesla is overfitting a lot to the west coast roads. The experience of driving in the US is so much better and safer than in Europe. Real chance that BYD will dominate the EU market

[Pedro F. Proença @flying_pedro]

Para um país socialista, o preço das autoestradas em Portugal é escandaloso. Na California nunca paguei um tostão para conduzir em highways de S. Diego a S. Francisco, de Los Angeles a Las Vegas.

[Pedro F. Proença @flying_pedro]

@levelsio This is true. Portuguese espresso sucks. Best I had were in the US.

[Pedro F. Proença @flying_pedro]

After 4 years at NASA-JPL, I decided to leave the US and move back to Portugal. The plan is to help Europe advance autonomous technology in critical areas and build its intelligent defense systems. Here’s some cool robots that kept me busy

[Pedro F. Proença @flying_pedro]

Protip for first time parents: Play Bob Marley to crying baby 🇯🇲

[Pedro F. Proença @flying_pedro]

I am among the 530 JPLers that were laid off today due to congress inaction. Around 17% of my division: Autonomous Systems was laid off! A lot of us are Foreign Nationals. This batch contains the most talented engineers I have met at JPL during my 4 yrs https://t.co/nGkw6zBoYm

[Pedro F. Proença @flying_pedro]

Plastic film from 🌮

[Pedro F. Proença @flying_pedro]

Game changing.

[Justin Johnson @jcjohnss]

I'm excited about Segment Anything released from FAIR today. It tackles an old problem (find objects in images) at large scale: trained on 11M images and 1B objects.
This is a new Foundation Model for Computer Vision - it recognizes any object in any context.

[Pedro F. Proença @flying_pedro]

I like this universe

[Pedro F. Proença @flying_pedro]

I never expected Mario Kart to prepare me for L.A. driving

[Pedro F. Proença @flying_pedro]

It's beer time

[Pedro F. Proença @flying_pedro]

It's Album Time
https://t.co/ueGXwgu3di