Paula Feldman

Behind the WiML Social at #ICLR2026, meet one of our organizers! ✨ @PauFeldman is a postdoc at @WeillCornell bridging generative AI & cardiovascular imaging, published at MICCAI and Medical Image Analysis. 📆 April 24 · 12-3 PM · Room 203C #WiML #WomenInML

Meet one of our panelists for the WiML Social at #ICLR2026! @AleksandraFaust, Director of Research at Google DeepMind, leads work on foundation and world models, with broad impact across AI, robotics, and healthcare. Join us on April 24 to hear her insights! #WiML

Join us at #ICLR2026 for the WiML Social✨ Panel: “There’s no single path: Navigating careers in academia, industry, and beyond”. Hear from amazing speakers on building careers across ML paths. 📆 Apr 24 | ⏰ 12–3 PM 📍 Room 203C #WiML #WomenInMachineLearning #MachineLearning

🚀 Excited to announce the WiML Social at @iclr_conf 2026! Join us for networking, career insights, and a great panel discussion. Open to all! 📆 April 24 ⏰ 12:00–3:00 PM 📍 Room 203 C, Convention Center #WiML #ICLR2026 #WomenInML #AI

@emmaiarussi 🫶🏻

Ayer se doctoró Paula. Cinco años muy intensos y también un poco complicados. Le tocó arrancar en plena pandemia, remoto desde Tucumán. Luego mudarse a Buenos Aires. Luego vivir de una beca doctoral Conicet. Aún así construyó una tesis con rigurosidad y belleza, en la intersección entre IA generativa, geometría y datos médicos. Ahora se va de posdoc a @cornell_tech en Nueva York. Es un orgullo enorme haberla acompañado ese camino.

@emmaiarussi @utditella @NeurIPSConf ☺️☺️

No saben el orgullo que me dan estos tres 😊 @PauFeldman, Martín Sinnona y Valentín Bonás del Laboratorio de IA en @utditella Vinieron a @NeurIPSConf #NeurIPS2025 (top conf. en IA) con tres trabajos increíbles 💪: • Paula: "VesselGPT: Autoregressive Modeling of Vascular Geometry" • Martín: "Do Large Language Models Understand Data Visualization Rules?" • Valentín: "De-rendering, Reasoning, and Repairing Charts with Vision-Language Models"

Day 1 at #MICCAI2025 ✅ @MICCAI_Society Amazing science and a taste of Korea in Daejeon! 🇰🇷 @PauFeldman will be presenting our 🤖 VesselGPT paper ✨ highlight ✨ on Thursday afternoon oral sessions: shorturl.at/pLKZz #MICCAI

VesselGPT produces diverse, anatomically realistic vessels, scales to deeper trees, and preserves key metrics like tortuosity. More at: github.com/LIA-DiTella/Ve…

Unlike prior work, VesselGPT models vessels beyond simple centerlines + radii. Our B-spline cross-sectional representation captures non-circular shapes, critical for realism.

We present an autoregressive model for vascular geometry that combines a VQ-VAE and GPT-2. The VQ-VAE learns a discrete vocabulary of vessel structures, and GPT-2 generates realistic geometries by sequencing this learned vocabulary.

Blood vessels 🩸 are complex, branching structures: hard to model, yet crucial for diagnosis, treatment planning, and simulations. Existing generative methods either oversimplify or fail to scale.

So proud of our PhD student @PauFeldman 🎉 She’ll be presenting “VesselGPT: Autoregressive Modeling of Vascular Geometry” at #MICCAI2025 @MICCAI_Society this September in Daejeon, South Korea 🇰🇷 — selected as 🚨 paper highlight & oral presentation! 🚨 More here 👇 conferences.miccai.org/2025/files/dow…

📢 Nuestro nuevo paper 📄 sobre síntesis de modelos de vasos sanguíneos 3D usando autoencoders variacionales ya está disponible en Medical Image Analysis 👉sciencedirect.com/science/articl… @PauFeldman @miguefainstein @vsiless @utditella

El Laboratorio de Inteligencia Artificial de la Escuela de Negocios #UTDT invita al Workshop in Machine Learning & Data Science edición 2024, en el que participarán especialistas del mundo académico y de la industria dedicados a la investigación en el área del Aprendizaje Automático y la Ciencia de Datos. Exponen: Ezequiel Álvarez (ICAS @unsamoficial), Cristian Antuña (@mutt_data), @PauFeldman (UTDT), @Sci_Brunobian (@liaa_icc), Michael Moreno (Quipu) y @melaniesclar (@LTIatCMU). Organizan: @RamiroHGalvez (UTDT) y @emmaiarussi (UTDT-@CONICETDialoga). Inscripción: bit.ly/4gioMpY

📢 ¡El próximo jueves 19/12 llega la 3ra edición de este Workshop en @utditella! Este año nos visitan Ezequiel Alvarez (UNSAM), @Brunobian (DC-UBA), @melaniesclar (UW/Meta) y más. 📅 Agenda completa e inscripción en el link 👇 shorturl.at/Cc3vJ

Not for the first time, all Nobel Prizes in scientific disciplines this year have been awarded to men

📢Journal Club is back! Title: VesselVAE: Recursive Variational Autoencoders for 3D Blood Vessel Synthesis Presented by: Ms Paula Feldman, Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina 📅 September 07th, at 4 pm CET

📢 último día para anotarse a este seminario especial de visión, gráficos e IA en @utditella 🧠👁️👾 👉 shorturl.at/Sb2yB

📢 Thrilled to be at #MICCAI2023 @MICCAI_Society showcasing our work from @utditella AI Lab on a deep learning generative framework for synthesizing 3D blood vessel geometry. 🎉 Kudos to the team @PauFeldman @vsiless!