VISUAL IMAGE PROCESSING LABORATORY
A research project on computational vision science.

Funded by Generalitat Valenciana under grant AICO-2020-136 (VIPLab) and CIAICO/2022/051 (IMaLeVICS) 

IMaLeVICS at IbPRIA 2025 in CoimbraDownload the presentation slides – PDFIMaLeVICS at IbPRIA 2025 in Coimbra

Jul 12, 2025 | Activity

IbPRIA 2025

The 12th Iberian Conference on Pattern Recognition and Image Analysis (IbPRIA 2025) took place from June 30 to July 3 in Coimbra, Portugal. Co-organized by the Portuguese APRP and the Spanish AERFAI chapters of the International Association for Pattern Recognition (IAPR), IbPRIA is a well-established forum for researchers and professionals working in pattern recognition, computer vision, and image analysis.

All accepted papers were published in the Springer Lecture Notes in Computer Science Series, and a selection of outstanding contributions will be invited to submit extended versions for high-impact journals. The conference also featured several awards, including Best Paper, Best Student Paper, and Ph.D. prizes from both national chapters.

Studying Just Noticeable Noise Differences

Our colleague Pedro Latorre Carmona presented “Masking of Gaussian noise in color images”, a collaborative study with Samuel Morillas, Rafael Huertas and Luis Miguel Calvo.

The work explores how the human visual system perceives Gaussian noise in digital color images through a psychophysical experiment involving 68 observers. By analyzing just-noticeable differences (JNDs) across different luminance levels and noise intensities, the team gained insight into how image quality is affected by sensor limitations.

The study uses forced-choice experiments and JOD predictors to refine noise perception models and suggests future lines of research in non-uniform regions and complex image contexts.

Vision, Noise and Perception

Understanding how we perceive noise is crucial for applications ranging from photography to medical imaging. At IMaLeVICS, we continue to combine rigorous experimentation with interpretable modelling to advance this understanding.

If you want to know more about it, contact us.