Slice priority prediction system for H.264 video

1. A system for prioritizing video slices of H.264 video bitstream comprising: a memory storage and a processing unit coupled to the memory storage, wherein the processing unit operates to execute a low complexity scheme to predict the expected cumulative mean squared error (CMSE) contributed by the loss of a slice of H.264 video bitstream, wherein the processing unit operates to execute a series of actions comprising the following steps performed in order:

assigning each slice a predicted value according to the low complexity scheme;

extracting video parameters during encoding process; and

using a generalized linear model to model CMSE as a linear combination of a plurality of the video parameters,

wherein the video parameters of step (b) are derived from analytical estimations using a Generalized Linear Model (GLM) over a video database, said database comprising videos of different characteristics, said characteristics comprising high and low motion, camera panning, zooming and still videos,

further comprising wherein the GLM is constructed in a training phase according to the following steps performed in order:

determining a distribution of the computed CMSE to be a Normal distribution with the Identity link function;

sequentially adding covariates using the forward selection technique whereby a best model is evaluated at each stage using the Akaike's Information Criterion (AIC);

generating regression coefficients during the training phase of the model;

validating a final model through a testing phase by predicting the CMSE for different video sequences, said different video sequences not in the training database; and

using the regression coefficients to predict expected CMSE values for each slice.

2. The system of claim 1 , wherein the extracted parameters of step (b) are: Motion Vectors (Motx and Moty), Residual Energy after motion compensation (MAXRESENGY), Initial Mean squared Error (IMSE), Initial Structural Similarity Index (IS SIM), Average Inter Partitions per macroblock (MB) (AvgInterParts), Error Propagation length (TMDR), slice type (slice_type) and signal characteristics such as signal mean and variance per slice (SigMean and SigVar).

3. The system of claim 2 , further comprising the step of examining the interactions between IMSE and MAXRESENGY and IMSE and TMDR.

4. The system of claim 1 , further comprising wherein the distortion prediction is computed by using a GLM model to predict CMSE, whereby CMSE is computed over the group of pictures (GOP) by dropping one H.264 video slice at a time, and further resolving the issue of GOP delay by developing a statistical model to predict the expected distortion.

5. The system of claim 1 , further comprising wherein per frame priority can be assigned where the slices of each frame are classified into two or four priorities separately in systems with stringent delay requirements.

6. The system of claim 1 , further comprising wherein the priorities can be assigned at the GOP level in delay tolerant systems.

7. A method of identifying the video parameters (from compressed H.264 bitstream) and their contribution to predicted CMSE by using the system of claim 1 .

8. A method of claim 7 , further comprising wherein there is no need to calculate CMSE over the entire GOP.

9. The method of claim 7 , wherein the encoder assigns priorities for each encoded slice based on its predicted CMSE without having to wait for the GOP statistics to assign the priorities unlike the traditional priority schemes.