Postdoc - Optimal color selection algorithms for multiplexed image inscription

falsifying one image would alter the others in a noticeable way. The development of this technology has resulted in three patents and to the publication of innovative scientific papers in selective journals. Three Hubert Curien Laboratory teams were involved in this development: Functional Surfaces and Materials, Image Science and Computer Vision, and Machine Learning. This demonstrates the interdisciplinary nature of the work.

Toppan Security and the LabHC are offering a one-year post-doctoral position renewable for an additional year to continue developing and optimizing this promising technology. We are seeking a postdoctoral researcher with a strong background in computational and algorithmic methods applied to physical systems. The successful candidate will contribute primarily to the development and improvement of image multiplexing algorithms by working with experimentally acquired color images of laser-processed nanostructured films. Close interaction with the experimental team is required to ensure the relevance, robustness, and validation of the proposed computational approaches.

Objectives

Over the past five years, LabHC has developed and implemented several algorithms to search for laser parameters that create the color combinations required for image multiplexing in experimental databases. The primary objective of this postdoctoral position is to enhance the methods employed for selecting laser parameters with a particular focus on the continuous multiplexing algorithm. The main step of this algorithm is restricting the full set of colors to the multiplexing solution, which is the set in which all possible color combinations in the different observation modes are possible. Currently, this is done by searching for the largest box that can be enclosed in the convex hull of the database. However, this box is not the optimal solution, so the first objective of the postdoc will be to automatically select the optimal polyhedron.

A second objective is to incorporate the properties of images to print into the algorithms. Constructing an imperfect multiplexing solution that better corresponds to the actual images could improve the color reproduction fidelity. Experimental work will be conducted to test the proposed modifications to our methods. They will consist of inscribing and characterizing the multiplexed images.

Requirements

Skills/Qualifications

• PhD in one of these fields: applied mathematics, applied physics, computational color science, or data science, with a demonstrated interest in applying computational methods to physical or experimental systems.
• Strong experience in scientific programming with Python, including developing algorithms and numerical methods.
• Proven experience in two or more of the following areas: multi-objective optimization, image processing, computational geometry (convex hulls, optimization in high-dimensional spaces), data-driven approaches applied to physical systems, ability to work with experimental datasets (including imperfect, limited, or noisy data) and to develop methods adapted to such constraints.
• Solid analytical skills and the ability to translate physical constraints and objectives into computational or algorithmic formulations.
• Basic familiarity with optics, photonics, or color science, or the demonstrated ability and motivation to rapidly acquire the necessary domain knowledge.
• Strong motivation for interdisciplinary research combining physics, algorithms, and applications.
• Good communication skills, and the ability to clearly document code, algorithms, and experimental results.
• Capacity to work in close collaboration with academic researchers from different disciplines and an industrial partner.
• Proactive attitude, creativity, and interest in exploring non-standard solutions.

Languages: English (Good)