Dataset structure

The user navigation dataset was recorded while 26 participants watched 150 looped frames of dynamic point cloud content in the Unity game engine. The location and orientation of their viewport is recorded at every rendered frame. All coordinates are based on the Unity world coordinate system. The XZ plane at the origin represents the floor. The point cloud dataset used in this work is:

Eugene d’Eon, Bob Harrison, Taos Myers, and Philip A. Chou. 2017. 8i Voxelized Full Bodies - A Voxelized Point Cloud Dataset, ISO/IEC JTC1/SC29 JointWG11/WG1 (MPEG/JPEG) input document WG11M40059/WG1M74006, Geneva.(January 2017)

The navigation dataset contains the following fields:

• FrameNumber: The frame number for the current point cloud frame being rendered
• HMDPX: The X coordinate of the viewport
• HMDPY: The Y coordinate of the viewport
• HMDPZ: The Z coordinate of the viewport
• HMDRX: The rotation of the viewport about the X axis
• HMDRY: The rotation of the viewport about the Y axis
• HMDRZ: The rotation of the viewport about the Z axis
• Participant: Identifier for the current participant
• Dataset: Identifier for the current point cloud sequence
• Viewframe: Frame number in the current session across multiple playback loops

The dataset identifier refers to the following sequences from the 8i voxelized full bodies dataset:

• H1: Longdress
• H2: Loot
• H3: Red and Black
• H4: Soldier

We kindly ask that should you use our dataset in your publication, that you would reference the following paper:

Shishir Subramanyam, Irene Viola, Alan Hanjalic, and Pablo Cesar. User Centered Adaptive Streaming of Dynamic Point Clouds with Low Complexity Tiling. ACM Multimedia (MM 2020), Seatle, United States. October 12-16, 2020