Most of tomorrow’s media formats tend to increase the sense of immersion into the content. Transforming the end-user/consumer’s experience by putting them into the middle of the action or information is the challenge, especially whenever narrative aspects are key.
The Advanced Media Content lab is building tools that enable the creation, storage, transportation, and sharing both for and by as many people as possible. It delivers expertise and technologies related to innovative dimensions of images (resolution, depth, color, light, etc.), sound (spatialized), and new rendering devices (screen, projection, AR/VR head-mounted displays with 3 or 6 degrees of freedom, etc.). It develops tools and solutions aimed at enabling the adoption of these future formats by industry professionals, and ultimately by end consumers. Holography and compression systems are key areas of expertise for the laboratory.
The challenges raised by new video and audio formats? Bringing even MORE realism and immersion to the experience, while still remaining easy to use and handle.
09.23.2022
Regroupement d'Activités de la Main Non-étiquetées
Nous proposons, dans cet article, une nouvelle approche qui utilise l'adaptation de domaine non supervisée (UDA) pour le regroupement d'activités de la main non étiquetées, c'est à dire le regroupement d'activités similaires dans des groupes distincts. L'approche proposée vise à exploiter les connaissances tirées des échantillons étiquetés du domaine source pour catégoriser les échantillons non…
read the publication
02.20.2023
Quality evaluation of the JPEG Pleno Holography Call for Proposals response
At its 93rd meeting, JPEG received the responses to the Call for Proposals on JPEG Pleno holography. This activity will lead to the first standardized coding solution for digital holograms. This paper describes the quality evaluation process, existing out of a subjective evaluation study and an objective evaluation based on preselected quality metrics. It also includes a critical evaluation of…
read the publication
05.24.2022
TwistSLAM: Constrained SLAM in Dynamic Environment
Classical visual simultaneous localization and mapping (SLAM) algorithms usually assume the environment to be rigid. This assumption limits the applicability of those algorithms as they are unable to accurately estimate the camera poses and world structure in real life scenes containing moving objects (e.g. cars, bikes, pedestrians, etc.). To tackle this issue, we propose TwistSLAM: a semantic,…
read the publication
05.16.2022
Metric Learning-Based Unsupervised Domain Adaptation for 3D Skeleton Hand Activities Categorization
First-person hand activity recognition plays a significant role in the computer vision field with various applications. Thanks to recent advances in depth sensors, several 3D skeleton-based hand activity recognition methods using supervised Deep Learning (DL) have been proposed, proven effective when a large amount of labeled data is available. However, the annotation of such data remains…
read the publication
2019
Product of the Year at the NAB Show in Las Vegas
2017
Technology Innovation Award at the NAB Show in Las Vegas
