Marc Serra Ortega

Research Engineer @ Computer Vision Center · Advanced MSc in AI @ KU Leuven

I work on document understanding and multimodal AI: vision-language models, document VQA, comics understanding, and efficient long-context inference. Vision and Language group at CVC, with Dimosthenis Karatzas.

alphaXiv Google Scholar LinkedIn
Computer VisionVision-LanguageMultimodal AIDocument AIDeep Learning

Papers

Linear-Time Context Compression with State-Space Encoding

Oct 2026 · Marc Serra Ortega, Andrey Barsky, Carlos Boned Riera, Tomás Ockier, Dimosthenis Karatzas

The Boa Compressor: a plug-and-play online context compressor that uses a state-space encoder to compress LLM context in linear time.

HF Paper arXiv alphaXiv

ICDAR2026 Competition on Multimodal Reasoning over Documents in Multiple Domains

Sep 2026 · Artemis Llabrés, Marc Serra Ortega, Tomàs Ockier, Samuel Ortega Cuadra, Amritpal Singh, Christos Georgakilas, Andrey Barsky, Ernest Valveny, Dimosthenis Karatzas

Competition report on reasoning-heavy document VQA across eight domains, from business reports and slides to maps, comics and engineering drawings.

HF Paper arXiv alphaXiv

CoSMo: A Multimodal Transformer for Page Stream Segmentation in Comic Books

Jul 2025 · Marc Serra Ortega, Emanuele Vivoli, Artemis Llabrés, Dimosthenis Karatzas

A multimodal Transformer for page stream segmentation in comics, with a new 20,800-page annotated dataset.

HF Paper arXiv alphaXiv

Experience

Education