ISMARPointCloudEyeTracking

The official repo for QAVA-DPC: Eye-Tracking Based Quality Assessment and Visual Attention Dataset for Dynamic Point Cloud in 6 DoF ISMAR 2023

Visual Saliency Map

The visual saliency maps generated per user can be downloaded from the following link:
👉 Download Visual Saliency Maps on Zenodo

File Naming Conventions

• Format: 001_A

  • 001: User ID
    
  • A: Session

• Format: H1_C2_R2_191

  • H1: Point cloud name
    
  • C2: Codec
    
  • R2: Distortion level
    
  • 191: Rotation degree

• Format: 4246452_rafa_084.txt

  • 4246452: Timestamp
    
  • rafa: Point cloud name
    
  • 084: Frame number

Label Definitions

• Point Cloud Names (H#):

  • H1: rafa2
    
  • H2: dancer
    
  • H3: exercise
    
  • H4: longdress
    
  • H5: soldier

• Codecs (C#):

  • C1: VPCC
    
  • C2: GPCC
    
  • C3: CWIPCL

Contents

The VisualSaliencyMap folder includes: - HeatValue: This subfolder contains the heat values for each frame in a dynamic point cloud sequence. Each point's heat value is saved in a text file, with values ranging from 0 to 1.

• HeatValuewithPointCloud: This subfolder provides visualizations of all heat values for each frame. The heat values are overlaid on top of the point cloud for each frame in all dynamic point cloud sequences.

Raw Gaze data for 40 users

In this folder, it includes all the experimental data related to the eye-tracking (in the json file) and the original opinion scores (in two txt files) of each user. It can be downloaded from: GazeData
user001 : useruserindex 001A.txt: userindexsession.txt
20230317-2301001A.json:dateuserindexsession.json

Calculated Quality Scores

You can find the calculated Mean Opinion Scores (Mos) and DMOS in the MOS/mos.csv and MOS/dmos.csv file.

Visualization:

This is the video of the H5C0R0_BackView

and H5C0R0_FrontView.

Quick Start

Device Specifications

• Processor Intel(R) Core(TM) i7-9700 CPU @ 3.00GHz 3.00 GHz
• Installed RAM 32,0 GB
• Device ID D415874E-183F-4E30-B8B7-FA373C373E84
• Product ID 00329-10333-35181-AA552
• System type 64-bit operating system, x64-based processor ## How to run it in Unity

Conditions of use

If you wish to use any of the provided material in your research, we kindly ask you to cite our paper. - BibTex @INPROCEEDINGS{10316522, author={Zhou, Xuemei and Viola, Irene and Alexiou, Evangelos and Jansen, Jack and Cesar, Pablo}, booktitle={2023 IEEE International Symposium on Mixed and Augmented Reality (ISMAR)}, title={QAVA-DPC: Eye-Tracking Based Quality Assessment and Visual Attention Dataset for Dynamic Point Cloud in 6 DoF}, year={2023}, volume={}, number={}, pages={69-78}, keywords={Point cloud compression;Measurement;Visualization;Solid modeling;Head-mounted displays;Gaze tracking;Inspection;Volumetric video;Dynamic point cloud;Visual saliency;Visual attention;Subjective quality assessment;Objective quality metrics;Eye tracking;6DoF}, doi={10.1109/ISMAR59233.2023.00021}}

About

The QAVA-DPC Dataset is maintained by the Distributed & Interactive Systems (DIS) research group at Centrum Wiskunde & Informatica (CWI).

Contact the authors - Xuemei Zhou: xuemei.zhou@cwi.nl