The IVC team has two psycho-visual test rooms in accordance with the international recommendations (ITU-R BT 500-11 for example). These recommendations allow to reproduce the tests all around the world like it have been done in Multi Media Test Plan of the Video Quality Expert Group (VQEG). In this test rooms, we realize some psycho-visual measurements experiment like quality evaluation on still images or on videos with different formats and resolutions. These evaluations can be done by expert or non-expert people.
Our plateform allow to measure different protocol and feedbacks about the contents under test. Some of them can be complementary to verify the results or to have another point of view. Some methodologies allow to measure the absolute or relative quality. Other methodologies allow to evaluate the perceptibility of an impairment. For example, in pair images, we can evaluated which one is the prefered image/video in term of quality, naturalness, luminosity point of view and so on. The power of the preference can also be evaluated.
Multi displays technology
Our multi display technology platform allow to measure the observer feedbacks with a protocol previously presented without any processing in the transmission channel. There are different sizes, technologies and connections to choose the better solution according to the scenario. This display diversity can be used to measure the impact of the display, with the realization of a test with several display.
Before each test, all our displays are calibrated .
Our multi-player platform allow to display images and videos from the QCIF (or less) resolution to Full HD resolution. The videos are played uncompressed (to manage what are displayed), in real-time up to Full HD (1920x1080 pixels) in progressive up to 60Hz. This platform also allow to record, in the same time to play, Full HD uncompressed videos in real-time. Its allowing to play, to process and to record, all in the same time, videos in real-time up to HD resolution up to 60Hz.
Multi resolution (from mobile resolution like QCIF to Full HD)
Multi analysis method
After each test, an analysis is done. This analysis is composed by the general computation like the Mean Opinion Score, the Confidence Interval and the outlier. Our analysis are completed by some statistical analysis to know if there is difference between two processings or to transform the scale of the result to allow the comparison with other methods and so on.
Realisation of subjective tests
The aims of the subjective test realization are :
- Academic collaborations. The academic collaborations can be done to evaluate a objective model on a specific database or to evaluate the difference between people around the world. Please see our publication with :
- Consortium projects. The IRCCyN laboratory/ IVC team participate to regional, national and international projects to provide its background in the image/video quality and to measure the efficiency of players, coders, displays or objective metrics with subjective tests, such as :
- Collaboration projects. Collaboration projects have been done with partners like :
- Internal works. It support the research of the IVC team on the psycho-visual domain.
We are open for collaboration. If you have some project, please contact us.