
Dunant
@DunantCR
Followers
28
Following
73
Media
0
Statuses
19
Ing. Mecatrónica - PUCP
Lima, Perú 🇵🇪
Joined September 2020
We keep moving 🚗
Stay tuned! This is how our #MappingVehicle a.k.a. the "ArtiMovil"can be spotted in the streets of Lima (from our early testing in Miraflores) -- integrating @MultiPacha LIDAR stream + our Drone&DashCam data as we work towards completing the map of Barranco this month!
0
0
3
🇵🇪 Mapping Barranco: [ 🟩🟩🟩🟥🟨⬜⬜⬜⬜⬜ ] (~30%): Here we show a second iteration of a sparse point-cloud computed from a medium set of drone flights (Trajectories shown in Red). Noticeable are the #PlazadeArmas (completed), the neighborhood between Av San Martin and Av Grau
0
6
18
8/ Will global leaders like Waymo, Wayve, WeRide or NVIDIA seize this opportunity, or will local innovators rise to the challenge? [Here's a modified Audi Add from the 90's conceptually expanding on this idea]
1
3
2
1/ “Lima York”: Is Latin America the next Frontier for Autonomous Driving? The autonomous driving industry is at a pivotal moment, shaped by geopolitics, regulation, and a race for high value data. A thread, https://t.co/z2Ygvhx9wq
7
10
12
📹We’ve released 200 driving videos from the Robusto-1 Dataset on @huggingface -- real, in-the-wild driving scenes from Lima, Cusco and Cajamarca in Peru 🇵🇪, now free to use! Use them to test VLMs, study edge cases, or fine-tune your own models #CVPR2025: https://t.co/N5bTsfkf0w
0
5
11
🤖 Thrilled to share: Our paper "Robusto-1 Dataset: Comparing Humans and VLMs on real out-of-distribution Autonomous Driving VQA from Peru" has been accepted at WDFM-AD workhop @CVPR 2025! How well do VLMs handle Peru’s real-world chaos? Watch the video & read more! 🎥 Link to
3
10
23
3/ Representational Similarity Analysis (Kriegeskorte et al. 2008) is a great tool to use for these experiments because it was designed to compare Artificial and Biological Systems via their internal representations or behaviour. We adapted it to answers of VQA for Humans VLMs
1
1
2
Can an AI chatbot understand driving as well as a human? 🚗💡 Our new paper explores this question using real data from Peru. 🇵🇪 Check it out! #AutonomousDriving
1/ How well do you think an AI chatbot could drive? Today we are excited to announce our first Self-Driving paper: “Robusto-1 Dataset: Comparing Humans and VLMs on real out-of-distribution Autonomous Driving VQA from Peru” 🇵🇪 #AI #AutonomousDriving
0
0
2
ROAM Challenge 1 (Adversarial Patch): https://t.co/ZMXZNRBePX ROAM Challenge 2 (Few-Shot Learning): https://t.co/jc4zS4k7HK Brought to you by @Artificio_Org + @UTECuniversidad + @huggingface at #ROAM @eccvconf
0
2
5
Today we're also excited to Announce the Parallel-Track Challenge 1: Adversarial Image (Patch) Attack for Autonomous Driving as part of the @roameccv2024 workshop at @eccvconf ! Submission link here:
huggingface.co
🥇 We're excited to Announce Challenge 2: Few-Shot Learning of Out-of-Distribution (OOD) objects for Autonomous Driving as part of the @roameccv2024 workshop at @eccvconf !
1
4
11
🥇 We're excited to Announce Challenge 2: Few-Shot Learning of Out-of-Distribution (OOD) objects for Autonomous Driving as part of the @roameccv2024 workshop at @eccvconf !
huggingface.co
1
9
35
Announcing the Challenge 2 Few-Shot Learning of Out-of-Distribution Challenge at the @roameccv2024 workshop! The Challenge will take place over 2 weeks! Sign-up & submit your entries now 🚀
🥇 We're excited to Announce Challenge 2: Few-Shot Learning of Out-of-Distribution (OOD) objects for Autonomous Driving as part of the @roameccv2024 workshop at @eccvconf !
0
3
7
1/ Excited to Announce a Series of Updates for the ROAM workshop 🌍 ! Our 2 Keynote Speakers will be Professor Fatma Güney (@ftm_guney) from @kocuniversity & Sergio Casas (@sergioksas) from @Waabi_ai & @UofT!
1
6
14
An example of running a Detectron on adversarial patches (virtual+real). In many cases the attack is not successful. This is a preview of Challenge1 @roameccv2024 where participants will have to submit patches that break a collection of private models for car+person detection.
0
3
5
🚀 Excited to announce the "ROAM: Robust, Out-of-Distribution, and Multi-Modal models for Autonomous Driving" workshop at #ECCV in co-organized between @Artificio_Org & @UTECuniversidad . 🌐
sites.google.com
The goal of the ROAM workshop is to introduce our new large-scale benchmark driving dataset, acquired from three diverse cities in South America (currently Peru), which provides a wide range of...
1
5
9
Today's a very special day as we celebrate @Artificio_Org 2nd birthday! It's been an epic ride so far, and we'll be sharing a lot of exciting news in May! 🤖🚘 For now, here's a Funny Birthday card from the A-team @Darofex @DunantCR @vfloresb21 @ArtDeza in Barranco,Lima, Peru🇵🇪
3
6
48
1/ Why has Autonomous Driving (AD) failed? We believe that the key to unlocking AD may be in 1) collecting rich out-of-distribution data from emerging economies such as #Peru; and 2) developing brain-aligned computer vision models #NeuroAI
https://t.co/jtWqX0pYJO
41
396
1K
Today we are excited to release our 2nd product and 1st API: Cortex Embeddings [🧠] . A Brain-Aligned Embedding that maps an Image into a Vector [ 🤖 ] to aid developers, scientists & businesses for any visual task [👁️]. https://t.co/KHRag0xsHW
0
17
37
Today we're excited to announce Seezlo : A Meta-Gallery and Search Engine platform for Visual Art 🎨 ; powered by our very own brain-aligned machine vision model🧠 [ that was not trained on Art! 🤯] https://t.co/AuTn0PDz41
6
82
669