Presentation at ACM-MM-19 TPC meeting: https://sites.google.com/comp.nus.edu.sg/mm19-tpc-meeting/workshop
With the growing popularity of X-Reality (VR/AR/MR) generation of global media-data is expected to accelerate at much higher rate. Most of these data need real-time processing such as real-time facial features/emotion recognition to environmental context recognition/mapping beyond the capabilities of an MR glass. The demand for high communication bandwidth and ultra-low latency (<15ms motion to photon latency) will drive the community to exploit the capabilities of the telco edge nodes and push telcos to build more powerful edge nodes. The real-time analytics of media data with strict real-time constraints will be the ‘killer application’ of 5G networks and edge computing. In addition, we believe, multimedia analytics and 5G/edge computing will be the key to push for mass adoption of XR devices. In this talk, we will explore some suitable media analytics methods and architecture to run media analytics with the 5G edge nodes.
2. Agenda
• VR/AR Consumer Adoption
• Role of Edge Nodes and 5G
• Key Challenges – Streaming,
Tracking, Analytics
3. Consumer Adoption-
Lack of quality contents
Any Killer-App for MASS
adoption?
Source: eMarketer.com
Lack of quality contents ranks high in a
recent survey on
“Primary Obstacle to Mass Adoption of
AR/VR Technology”
4. Consumer Adoption – Not compact enough
• Does not look and feel like everyday glasses to use for most of the
context sensitive applications.
best AR glass on the market
5. Consumer Adoption – other issues
• VR Hardware Limitations
• Motion Sickness with
controller based navigation,
FoV (110 vs required 190),
Resolution
• AR
• Lack of a geospatial social
“AR Cloud” that would
make the glasses
contextually aware
Field of View (FoV)
Eye 350 + Head 600 = 950. Both sides : 95 x 2 = 1900
Technology is still not perfect for consumer/entertainment market
6. Consumer Adoption – Focus shifts towards
Commercial/Industrial market
Technology is still not perfect for consumer/entertainment market
8. Technologies to enable MASS adoption
(Making the standalone devices powerful with the
edge nodes)
1. VR Cloud/edge Streaming
2. Edge nodes & analytics
•Object Recognition
•Realtime 3D Mesh Reconstruction and Crowdsourcing
to Build the world (AR Cloud)
11. Streaming Requirements for Immersion
• Low latency
• 15 ms for VR. [some studies state 20ms]
•1/90s frame time [~ 11 ms]
•Bandwidth
•Current: 4K Video
•Future: 8K Video and 6 DoF
17. Optimising Gamelet for VR
EncodingDecoding
@ Edge Nodes
@ VR Headset
VR Rendering Pipeline with Streaming
e to e latency < 20ms
Independent Hardware for
Rendering and Encoding
18. Source: Luyang Liu, MobiCom
Optimising Gamelet
for VR
Parallelizing
Rendering and
Streaming
19. Source: Luyang Liu, MobiCom
Optimising Gamelet for VR Parallelizing Rendering and
Streaming
20. Source: Luyang Liu, MobiCom
Optimising Gamelet for VR Parallelizing Rendering and
Streaming
22. Realtime 3D Mesh
Reconstruction and
Crowdsourcing to Build the
world mesh with AR data
(“Cloud AR”)
• A challenging process with resource limited
glasses/phones
23. Realtime 3D Mesh Reconstruction and
Crowdsourcing to Build the world mesh (Cloud AR)
28. Cloud vs 5G-Edge: Analytics [Real-time]
Real-time Analytics can be cone in Cloud, 5G-Edge, or Client.
Realtime Video Analytics at the Edge
(Streaming & Analytics)
Options & Trade-offs:
• Cloud Only
• 5G-Edge Only
• Client Node Only
• Cloud and 5G-Edge
• 5G-Edge and Client
• Cloud, 5G-Edge and Client