LFDD: Light field image dataset for performance evaluation of objective quality metrics
An increase in research activity around plenoptic content can be seen in recent years. As the communities around the different modalities grow, so does the demand for publicly available annotated datasets of suitable content. The datasets can be used for a multitude of purposes, such as to design novel compression algorithms,subjective evaluation methodologies or objective quality metrics. In this work, a new publicly available annotated light field image dataset is presented. The dataset consists of scenes corrupted by state-of-the-art image and video compression algorithms (JPEG, JPEG 2000, BPG, VP9, AV1, AVC, HEVC), noise, and geometric distortion. For the subjective evaluation of the included scenes, a modified version of the Double Stimulus Impairment Scale (DSIS) methodology was adopted. The views of each scene were organized in a pseudo-sequence and played to the observers as animations. The resulting objective and subjective scores, together with additional data, are included in the dataset. The dataset can be used to evaluate the performance of currently used visual quality metrics as well as for the design of new ones.
The individual scenes of the dataset are available below. The whole dataset can be downloaded here:
Cross-Content Evaluations in the Subjective Quality Assessment of Light Field Images
Additional changes to the contents below might be done upon the approval of the accompanying paper.
Update 1.1
This update is focused on the differences between the results of a Double Stimulus Impairment Scale (DSIS) methodology (the original subjective evaluation) and Pairwise Comparison (PC) methodology. The update is composed of an additional scene (the Herbs scene) and additional subjective scores, including cross-content evaluations. The scenes included in the update were corrupted only by state-of-the-art image and video compression algorithms (JPEG, JPEG 2000, BPG, VP9, AV1, AVC, HEVC).
