O
Olivier Deforges
Researcher at University of Rennes
Publications - 226
Citations - 2019
Olivier Deforges is an academic researcher from University of Rennes. The author has contributed to research in topics: Data compression & Image compression. The author has an hindex of 20, co-authored 216 publications receiving 1518 citations. Previous affiliations of Olivier Deforges include Intelligence and National Security Alliance & Centre national de la recherche scientifique.
Papers
More filters
Journal ArticleDOI
NIQSV+: A No-Reference Synthesized View Quality Assessment Metric
TL;DR: This paper proposes a novel no-reference image quality assessment method for 3-D synthesized views (called NIQSV+), which can evaluate the quality of synthesizer views by measuring the typical synthesis distortions: blurry regions, black holes, and stretching, with access to neither the reference image nor the depth map.
Proceedings ArticleDOI
Salgan360: Visual Saliency Prediction On 360 Degree Images With Generative Adversarial Networks
TL;DR: The SalGAN, a 2D saliency model based on the generative adversarial network, is extended to SalGAN360 by fine tuning the SalGAN with a new loss function to predict both global and local saliency maps.
Journal ArticleDOI
Color LAR Codec: A Color Image Representation and Compression Scheme Based on Local Resolution Adjustment and Self-Extracting Region Representation
TL;DR: An efficient content-based image coding called locally adaptive resolution (LAR) offering advanced scalability at different semantic levels, i.e., pixel, block, and region, is presented, which provides a representation at a region level while avoiding any contour encoding overhead.
Journal ArticleDOI
Real-Time Selective Video Encryption based on the Chaos System in Scalable HEVC Extension
TL;DR: A real-time selective video encryption solution in the scalable extension of High Effciency Video Coding (HEVC) standard, referred to as SHVC, that encrypts a set of sensitive SHVC parameters with a minimum delay and complexity overheads and preserves all SHVC functionalities.
Journal ArticleDOI
Fast recursive grayscale morphology operators: from the algorithm to the pipeline architecture
TL;DR: A new algorithm for efficient computation of morphological operations for gray images and the specific hardware based on a new recursive morphological decomposition method of 8-convex structuring elements by only causal two-pixelstructuring elements (2PSE).