Manel Baradad Jurjo
Machine learning and computer vision researcher specializing in pre- and post-training
for generative vision systems. Experience leading multimodal diffusion and image-editing
models from data curation and distributed training through evaluation and production deployment.
PhD from MIT CSAIL with research spanning generative models, representation learning, and 3D vision.
TECHNICAL EXPERTISE Generative vision · Foundation models ·
Pre-training · Post-training / fine-tuning · Diffusion models ·
ControlNet · LoRA · Image editing · Vision-language models ·
LLMs / RAG · PyTorch · Distributed training · Data curation · Model evaluation
Professional & Research Experience
Amazon Science, BCN, Spain
Sep 2024 – Present
Applied Scientist
- Led the development and pre- and post-training of multimodal diffusion
and image-editing foundation models for product image generation, placement, and marketing creatives
deployed on the Amazon platform.
- Developed controllable generation and efficient adaptation methods using ControlNet
models and LoRA adapters, incorporating creative-feedback signals to improve product fidelity, composition,
and instruction adherence.
- Built end-to-end PyTorch pipelines spanning multimodal data curation,
distributed multi-node, multi-GPU training on a mid-scale compute cluster,
model evaluation, and production deployment.
- Amazon AGI: designed and trained neural cartridges for
retrieval-augmented generation (RAG), targeting context compression and efficient
long-context inference for large language models.
Google, GeometricAI, NYC
Jun 2023 – Feb 2024
Research Internship
Researched wavelet-based volumetric rendering.
Google, Viscam, Cambridge, USA
Jun 2022 – Sep 2022
Research Internship
Researched object-centric depth reconstruction from a single image.
Uber ATG, San Francisco (remote)
Jun 2020 – Sep 2020
Research Summer Internship
Researched novel-view synthesis for autonomous-driving simulation.
Earlier Experience
Research Assistant, MIT CSAIL (2017–2018) · Research Assistant, Image Processing Group,
UPC (2016–2017) · Software Engineer, Zhilabs (2015–2016) · Research Assistant,
Algorithmics and Learning Lab, UPC (2015)
Education
CSAIL, Massachusetts Institute of Technology (MIT)
2018 – 2024
PhD, Computer Science
Advisor: Antonio Torralba
Polytechnic University of Catalonia (UPC)
2015 – 2017
MS, Electrical Engineering
CFIS, UPC
2011 – 2015
BS, Electrical Engineering · BS, Computer Engineering
Selected Publications
PLACID: Identity-Preserving Multi-Object Compositing via Video Diffusion
with Synthetic Trajectories
Gemma Canet Tarrés, Manel Baradad, Francesc Moreno-Noguer, Yumeng Li.
CVPR 2026
Separating Knowledge and Perception with Procedural Data
Adrián Rodríguez-Muñoz, Manel Baradad, Phillip Isola, Antonio Torralba.
ICML 2025
A Vision Check-up for Language Models
Pratyusha Sharma, Tamar Rott Shaham, Manel Baradad, Stephanie Fu,
Adrián Rodríguez-Muñoz, Shivam Duggal, Phillip Isola, Antonio Torralba.
CVPR 2024 (Highlight)
Procedural Image Programs for Representation Learning
Manel Baradad, Chun-Fu Chen, Jonas Wulff, Tongzhou Wang,
Rogerio Feris, Antonio Torralba, Phillip Isola.
NeurIPS 2022
Learning to See by Looking at Noise
Manel Baradad*, Jonas Wulff*, Tongzhou Wang, Phillip Isola, Antonio Torralba.
NeurIPS 2021 (Spotlight)
Height and Uprightness Invariance for 3D Prediction From a Single View
Manel Baradad, Antonio Torralba.
CVPR 2020
Inferring Light Fields From Shadows
Manel Baradad, Vickie Ye, Adam B. Yedidia, Frédo Durand,
William T. Freeman, Gregory W. Wornell, Antonio Torralba.
CVPR 2018 (Spotlight)
Background Prompting for Improved Object Depth
Manel Baradad, Yuanzhen Li, Forrester Cole, Michael Rubinstein,
Antonio Torralba, William T. Freeman, Varun Jampani.
arXiv 2023
Honors and Awards
"laCaixa" Fellowship (2018; two years of graduate tuition and stipend) ·
Fundació Catalunya-La Pedrera Fellowship (2015; MS tuition) ·
CFIS Fellowship (2012; BS tuition)
Languages English: Fluent · Spanish, Catalan: Native
· French: Beginner