SV-QoE: Perceptual Quality Assessment of Spatial Videos on Apple Vision Pro
The SV-QoE dataset consists of stereoscopic and 2D video sequences generated entirely inside Unity Engine (v.6000.0.27f1). The virtual cameras were calibrated to simulate human binocular vision:
- Stereo baseline: 65 mm (average human interpupillary distance)
-
Left/Right camera positions:
- Left_Camera X = −0.325 m, Y = 0, Z = 0
- Right_Camera X = +0.325 m, Y = 0, Z = 0
- Central_Camera X = 0, Y = 0, Z = 0
- Field of View (horizontal): 137°, providing a wide-angle view.
- Each camera rendered at 4K resolution (3840 × 2160) with 60 FPS output.
-
Rendering Settings:
- Clipping planes: Near = 0.3 units, Far = 1000 units
- Depth priority: -1
- Occlusion culling enabled for performance
- HDR and MSAA enabled
- Render Path: set to “Use Graphics Settings”
Download Dataset
Playback Devices
Apple Vision Pro
Immersive spatial video playback experience
Meta Quest 3
Compatible VR headset for spatial video viewing
BibTeX Citation:
@inproceedings{gholam02025sv-qoe,
title={Perceptual Quality Assessment of Spatial Videos on Apple Vision Pro},
author={Gholami, Afshin and Baldoni, Sara and Battisti, Federica and Zhou, Wei and Timmerer, Christian and Amirpour, Hadi},
year = {2025},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3746269.3760422},
doi = {10.1145/3746269.3760422},
booktitle = {Proceedings of the 3nd International Workshop on Interactive EXtended Reality},
location = {Dublin, Ireland}
}Contact
Get in touch with us for any questions or collaborations.
Email: Hadi.Amirpour@aau.at
Licence
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.