Research areas

Research areas

    • Analysis and prediction of visual appearance

      Visual appearance plays a central role in many domains especially in (computational) printing, art design and visual diagnostics. An objective approach able to combine photometry, image processing and machine learning techniques makes possible the prediction of appearance attributes as color, texture, glossy and perceptual transparency from a reduced amount of data. We develop physical/digital models comprising light flux propagation and scattering, reflectance estimation from RGB images, visual saliency.

      Read more

    • Analysis of motion, human expression and behavior

      In this topic, we specifically address tasks related to (i) motion analysis where video sequences are processed to estimate velocity or track moving objects or particles, (ii) human expression and behavior understanding which includes detection and recognition of face expressions, micro-expressions, gestures and specific events or actions related to human activities. The originality of our approach is to process heterogeneous data as signal, image, depth, video...

      Read more

    • Description and representation of images

      In computer vision, to address high level semantic tasks as image recognition or scene analysis, modern approaches rely on new methods to describe and represent the visual information. In this topic we develop full representation models and algorithms including feature detection, local descriptions, statistical representation, machine learning, deep networks. We consider color, RGBD or infrared images or videos and address applications as health, digital technologies or cultural heritage.

      Read more

    • Unconventional imaging

      Combining imaging techniques and reconstruction algorithms extends the capabilities of conventional imaging. By working from optical system design to experimental set-ups, we can improve imaging for large telescopes, develop color lensless microscopy for microbiology, fluorescence imaging for ophthalmology and spectral imaging for color appearance. Thanks to an inverse problems approach, we can achieve a better reconstruction in digital holography and an improved image restauration in radar teledetection. A special focus is placed on the co-design of both the optical system and the image processing.

      Read more