Title: Channel-Adaptive Video Streaming Using Packet Path Diversity and Rate-Distortion Optimized Reference Picture Selection
1Channel-Adaptive Video Streaming Using Packet
Path Diversity andRate-Distortion Optimized
Reference Picture Selection
2Motivation
Content Server
- Channel-adaptive packet dependency control
- Higher error-resilience no packet retransmission
- Low-latency streaming (lt1 second) for interactive
applications - Liang, Flierl, Girod, ICIP02
- Path diversity
- Uncorrelated channel statistics
- Mitigate negative impact on data
- This work
- Low-latency streaming using multiple paths for
further improvement - R-D optimization for multiple streams
Data traffic
Data traffic
Relay Server
1
Relay Server
2
Client
3Outline
- Prior work
- Channel-adaptive reference picture selection
- Optimal packet dependency control for multiple
streams - Results and video demo
4Prior Schemes
- Video redundancy coding (VRC) H.263
- Multiple state coding
- Apostolopoulos
- RPS-NACK
- Lin et. al., ICME01
1
3
Path 1
S
S
2
4
Path 2
Path 1
0
2
8
4
6
10
1
3
5
7
9
Path 2
5Dependency Control over Multiple Paths
Path 1
Path 2
n-5 n-4 n-3 n-2
n-1 n
Long-Term Memory V5
Prediction modes v1, 2, V, I
6Path Selection
- Path selection to minimize the distortion of
next frame and take advantage of path diversity - Alternate when both channels are good
- Send small probe packets over the channel in bad
state
7Results (1)
Channel loss_rate_1 loss_rate_2 15 LTM
len12 Avg burst len8 Feedback delay6
8Results (2)
Channel loss_rate_1 loss_rate_2 15 LTM
len12 Avg burst len8 Feedback delay6
9Video Demo (1)
Proposed scheme RPS-NACK Mother-Daughter 318
Kbps 20 loss
10Video Demo (2)
Proposed scheme RPS-NACK Foreman 185
Kbps 20 loss
11Conclusions
- Eliminated the need for retransmission
- low-latency (lt1 second) streaming over
multi-path - Increased error-resilience by using
- Packet path diversity
- Feedback-based path selection
- Optimal packet dependency control
- Improved tradeoff between error-resilience
(independent streams) and coding efficiency using
R-D optimization
12Estimate of the Distortion
n-3 n-2 n-1 n
q
q
p
q
p
p