Cross-codec encoding optimizations for video transcoding

Abstract

A method for sharing the motion estimation and mode decision results and decisions of one codec with another codec is disclosed. A video is received to be transcoded into a plurality of different output encodings of a plurality of different codecs. Each codec has a different video encoding format. A shared motion estimation and a shared mode decision processing of the video are performed. One or more results of the shared mode decision processing shared across the plurality of different codecs are used to encode the video into the plurality of different output encodings of the plurality of different codecs.

What is claimed is: 
     
         1 . A method, comprising:
 receiving a video to be transcoded into a plurality of different output encodings of a plurality of different codecs, wherein each codec has a different video encoding format;   performing a shared mode decision processing of the video; and   using one or more results of the shared mode decision processing shared across the plurality of different codecs to encode the video into the plurality of different output encodings of the plurality of different codecs, comprising:
 mapping at least one result of the shared mode decision processing that is not compatible with a first codec of the plurality of different codecs to a mapped mode decision processing result compatible with the first codec of the plurality of different codecs. 
   
     
     
         2 . The method of  claim 1 , wherein the shared mode decision processing is not compatible with the first codec of the plurality of different codecs with a first video encoding format, and wherein the shared mode decision processing is compatible with a second codec of the plurality of different codecs with a second video encoding format. 
     
     
         3 . The method of  claim 2 , further comprising sending the at least one result of the shared mode decision processing to a mapping module that maps the at least one result of the shared mode decision processing to the mapped mode decision processing result compatible with the first video encoding format. 
     
     
         4 . The method of  claim 3 , further comprising sending the mapped mode decision processing result to a first standard-specific module compatible with the first video encoding format for further processing, wherein the first standard-specific module comprises a first standard-specific final encoding module and a first standard-specific filter engine. 
     
     
         5 . The method of  claim 4 , wherein the first standard-specific final encoding module performs prediction, transform and quantization, and entropy coding. 
     
     
         6 . The method of  claim 3 , wherein the mapping module maps the at least one result of the shared mode decision processing to the mapped mode decision processing result using a mapping function based on a machine learning model, wherein the machine learning model is trained based on mode decisions collected on one or more video decoders. 
     
     
         7 . The method of  claim 3 , wherein the mapping module performs additional mode decision processing, wherein the mapped mode decision processing result is based on the at least one result of the shared mode decision processing and results of the additional mode decision processing. 
     
     
         8 . The method of  claim 3 , wherein the at least one result of the shared mode decision processing comprises intermediate results of mode decision processing, and wherein the mapping module performs additional mode decision processing based on the intermediate results of mode decision processing to form the mapped mode decision processing result. 
     
     
         9 . The method of  claim 3 , wherein the mapping module uses the at least one result of the shared mode decision processing as one or more initial values for a refinement search for one or more refined mode decision processing results to form the mapped mode decision processing result. 
     
     
         10 . The method of  claim 2 , further comprising directly sending the at least one result of the shared mode decision processing to a second standard-specific module compatible with the second video encoding format for further processing, wherein the second standard-specific module comprises a second standard-specific final encoding module and a second standard-specific filter engine. 
     
     
         11 . The method of  claim 10 , wherein the second standard-specific final encoding module performs prediction, transform and quantization, and entropy coding. 
     
     
         12 . A system, comprising:
 an interface configured to receive a video to be transcoded into a plurality of different output encodings of a plurality of different codecs, wherein each codec has a different video encoding format; and   a processor coupled to the interface and configured to:
 perform a shared mode decision processing of the video; and 
 use one or more results of the shared mode decision processing shared across the plurality of different codecs to encode the video into the plurality of different output encodings of the plurality of different codecs, comprising: 
 mapping at least one result of the shared mode decision processing that is not compatible with a first codec of the plurality of different codecs to a mapped mode decision processing result compatible with the first codec of the plurality of different codecs. 
   
     
     
         13 . The system of  claim 12 , wherein the shared mode decision processing is not compatible with the first codec of the plurality of different codecs with a first video encoding format, and wherein the shared mode decision processing is compatible with a second codec of the plurality of different codecs with a second video encoding format. 
     
     
         14 . The system of  claim 13 , wherein the processor is configured to process the mapped mode decision processing result using a first standard-specific final encoding module and a first standard-specific filter engine, wherein the first standard-specific final encoding module performs prediction, transform and quantization, and entropy coding. 
     
     
         15 . The system of  claim 13 , wherein the processor is configured to map the at least one result is of the shared mode decision processing to the mapped mode decision processing result using a mapping function based on a machine learning model, wherein the machine learning model is trained based on mode decisions collected on one or more video decoders. 
     
     
         16 . The system of  claim 13 , wherein the processor is configured to map the at least one result of the shared mode decision processing to the mapped mode decision processing result by performing additional mode decision processing, wherein the mapped mode decision processing result is based on the at least one result of the shared mode decision processing and results of the additional mode decision processing. 
     
     
         17 . The system of  claim 13 , wherein the at least one result of the shared mode decision processing comprises intermediate results of mode decision processing, and wherein the processor is configured to map the at least one result of the shared mode decision processing to the mapped mode decision processing result by performing additional mode decision processing based on the intermediate results of mode decision processing to form the mapped mode decision processing result. 
     
     
         18 . A system, comprising:
 an interface receiving a video to be transcoded into a plurality of different output encodings of a plurality of different codecs, wherein each codec has a different video encoding format;   a mode decision module configured to perform a shared mode decision processing of the video; and   a mapping module configured to use one or more results of the shared mode decision processing shared across the plurality of different codecs to encode the video into the plurality of different output encodings of the plurality of different codecs, comprising:
 mapping at least one result of the shared mode decision processing that is not compatible with a first codec of the plurality of different codecs to a mapped mode decision processing result compatible with the first codec of the plurality of different codecs. 
   
     
     
         19 . The system of  claim 18 , wherein the shared mode decision processing is not compatible with the first codec of the plurality of different codecs with a first video encoding format, and wherein the shared mode decision processing is compatible with a second codec of the plurality of different codecs with a second video encoding format. 
     
     
         20 . The system of  claim 19 , wherein the mapping module uses the at least one result of the shared mode decision processing as one or more initial values for a refinement search for one or more refined mode decision processing results to form the mapped mode decision processing result.