Patrick Le Callet

Pseudo No Reference image quality metric using perceptual data hiding

Regarding the important constraints due to subjective quality assessment, objective image quality assessment has recently been extensively studied. Such metrics are usually of three kinds, they might be Full Reference (FR), Reduced Reference (RR) or No Reference (NR) metrics. We focus here on a new technique, which recently appeared in quality assessment context: data-hiding-based image quality metric. Regarding the amount of data to be transmitted for quality assessment purpose, watermarking based techniques are considered as pseudo noreference metric: A little overhead due to the embedded watermark is added to the image. Unlike most existing techniques, the proposed embedding method exploits an advanced perceptual model in order to optimize both the data embedding and extraction. A perceptually weighted watermark is embedded into the host image, and an evaluation of this watermark allows to assess the host images quality. In such context, the watermark robustness is crucial; it must be suffciently robust to be detected after very strong distortions, but it must also be suffciently fragile to be degraded along with the host image. In other words, the watermark distortion must be proportional to the images distortion. Our work is compared to existing standard RR and NR metrics in terms of both the correlation with subjective assessment and of data overhead induced by the mark.

From low-level perception to high-level perception: a coherent approach for visual attention modeling

The saliency-based or bottom-up model of visual attention presented in this paper deals with still color images. The model we built is based on numerous properties of the human visual system (HVS), thus providing a biologically plausible system. The computation of early visual features such as color and orientation is a key step for any bottom-up model and the way to extract these visual features easily permits to differentiate a model from an other. The novelty of the proposed approach lies on the fact that the computation of early visual features is fully based on a HVS model consisting in projecting the picture into an opponent-colors space, applying a perceptual decomposition, contrast sensitivity and masking functions. Moreover, a strategy essentially based on a center surround mechanism and on the perceptual grouping phenomena underscores conspicuous locations by combining visual feature maps. A saliency map which is defined as a 2D topographic representation of conspicuity is then deduced. The model is applied to a number of natural images. Our results are then compared with the results of a well-know bottom-up model.

Subjective assessment methodology for preference of experience in 3DTV

The measurement of the Quality of Experience (QoE) in 3DTV recently became an important research topic as it relates to the development of the 3D industry. Pair comparison is a reliable method as it is easier for the observers to provide their preference on a pair rather than give an absolute scale value to a stimulus. The QoE measured by pair comparison is thus called “Preference of Experience (PoE)”. In this paper, we introduce some efficient designs for pair comparison which can reduce the number of comparisons. The constraints of the presentation order of the stimuli in pair comparison test are listed. Finally, some analysis methods for pair comparison data are provided accompanied with some examples from the studies of the measurement of PoE.

A human visual model-based approach of the visual attention and performance evaluation

In this paper, a coherent computational model of visual selective attention for color pictures is described and its performances are precisely evaluated. The model based on some important behaviours of the human visual system is composed of four parts: visibility, perception, perceptual grouping and saliency map construction. This paper focuses mainly on its performances assessment by achieving extended subjective and objective comparisons with real fixation points captured by an eye-tracking system used by the observers in a task-free viewing mode. From the knowledge of the ground truth, qualitatively and quantitatively comparisons have been made in terms of the measurement of the linear correlation coefficient (CC) and of the Kulback Liebler divergence (KL). On a set of 10 natural color images, the results show that the linear correlation coefficient and the Kullback Leibler divergence are of about 0.71 and 0.46, respectively. CC and Kl measures with this model are respectively improved by about 4% and 7% compared to the best model proposed by L.Itti. Moreover, by comparing the ability of our model to predict eye movements produced by an average observer, we can conclude that our model succeeds quite well in predicting the spatial locations of the most important areas of the image content.