AI-FUSE (AI For Universal Scene Understanding: from cells to ecosystems) main goal is to improve key aspects of scene understanding algorithms, namely generalization capabilities, efficiency, robustness and improved explainability of solutions and evaluations. The AI-FUSE interdisciplinary team will devise new solutions with broader applicability in real world applications than the existing methods. Our objectives include scientific advances in key AI techniques as well as in different and diverse application domains that can make use of these techniques: robotics deployment in human environments, environment monitoring and immunology research.
Embedded Files
This coordinated project aims to push the frontiers in the AI field by investigating and developing novel universal scene understanding AI algorithms and benchmarks that can democratize AI adoption and usage. In particular, AI-FUSE will develop novel algorithms for video understanding in different application scenarios of interest in many domains, from the private sector (surveillance or robotics) to public institutions (health and ecology). The project's ambition is not to tailor the algorithms for each application but to create a general algorithmic workflow for all and any domains, so that the proposed approaches can be applied broadly, including fields with currently low AI adoption.
To contribute to a broader adoption of AI in different fields, during the project the AI-FUSE team will first thoroughly study the interdisciplinary requirements from the team's diverse backgrounds, to consider among other aspects, the possible synergies. We will develop new algorithms, models and evaluation frameworks that will boost the applicability of scene understanding techniques. AI-FUSE will investigate more universal solutions, thanks to the diverse viewpoints, and explainable, thanks again in part to the inclusion of interdisciplinary ideas from the design.
AI-FUSE will innovate in the AI field, with novel video understanding approaches, including technical contributions to improve their efficiency, make them more effective and facilitate their applicability in natural scenarios related to disciplines where AI is not broadly adopted yet, such as cell biology or ecology.
PARTICIPANTS
PARTICIPANTS
Universidad
de Zaragoza
Universidad
de Zaragoza
PIs and Coordinators:
Ana C. Murillo
Eduardo Montijano
PIs and Coordinators:
Ana C. Murillo
Eduardo Montijano
Team:
Rubén Martínez-Cantín
Luis Montesano
Luis Riazuelo
Darío Suarez
Cesar Borja
Nerea Gallego
David Morilla
Fernando Peña
Carlos Plou
Team:
Rubén Martínez-Cantín
Luis Montesano
Luis Riazuelo
Darío Suarez
Cesar Borja
Nerea Gallego
David Morilla
Fernando Peña
Carlos Plou
Universidad
Pablo Olavide
Universidad
Pablo Olavide
PIs:
Luis Merino
Jesús Capitán
PIs:
Luis Merino
Jesús Capitán
Team:
Fernando Caballero
Pablo Pueyo
Team:
Fernando Caballero
Pablo Pueyo
Fundación para la investigación médica aplicada
Fundación para la investigación médica aplicada
PIs:
Carlos Ortiz de Solorzano
Alvaro Teijera Sánchez
PIs:
Carlos Ortiz de Solorzano
Alvaro Teijera Sánchez
Team:
María Sandra Hervás Stubbs
Ines Antomas Navarro
Pedro Castillo Rosique
Iván Cortés Domínguez
Tomás Muñoz Santoro
Carlos Sobejano de la Merced
Team:
María Sandra Hervás Stubbs
Ines Antomas Navarro
Pedro Castillo Rosique
Iván Cortés Domínguez
Tomás Muñoz Santoro
Carlos Sobejano de la Merced
Page updated
Report abuse