Coloma Ballester

Coloma Ballester is Full Professor of Applied Mathematics at the Department of Engineering of the Universitat Pompeu Fabra in Barcelona, where she coordinates the Intelligent Multimodal Vision Analysis (IMVA; previously IPCV image processing and computer vision) research group and the Interactive Digital Image (IDI) research unit.

Her research has been devoted to the development of models and algorithms for image processing and computer vision and to their mathematical analysis. This has led to pioneering publications in variational methods, geometric models, partial differential equations, non-local approaches and data-driven learning strategies for imaging, and to contributions to several problems: image inpainting, interpolation and disocclusion, image segmentation, shape from texture, object recognition, mathematical morphology including data structures, image fusion and super-resolution, restoration and denoising, depth estimation, video interpolation for slow camera motion and temporal super-resolution, optical flow and occlusion detection, motion inpainting, de-interlacing, colorization, out-of-distribution and anomaly detection, non-local methods and its applications, visual perception and its modeling, mathematical analysis of the TV flow, generative models and its use for solving inverse problems, self-supervised models for the geometric and semantic analysis of visual data and applications.

She currently works on theory and applications for the automatic understanding of visual content, sometimes involving multiple modalities including natural language, audio, and of course vision. She is interested in both model-based and learning strategies for multimodal scene understanding, together with its mathematical analysis.

Link to some of her recent research contributions in specific contexts within automatic scene analysis and understanding, including self-supervised video summarization and highlights, video understanding and editing tools for media broadcasting, multi-view restoration and editing, sign language translation and generation, film and visual arts analysis and understanding, sports video analysis.