Video Random Access Unit Types for Mid-Bitstream CRA Pictures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video codec standards limit the flexibility of random access points in video bitstreams, restricting the ability to efficiently perform operations like fast forwarding, rewinding, and seamless switching between video streams.
Innovation Solution
Introduce clean random access (CRA) pictures that can be placed mid-bitstream and allow for potentially non-decodable leading pictures, using flags or indicators to manage these non-decodable pictures, and define new unit types for RAP pictures to simplify mapping to container formats.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional video codec standards are used with limited random access points, then decoding complexity is reduced, but the flexibility and adaptability of video delivery and editing operations deteriorate
Solution Approach 1:
The video bitstream is segmented into multiple access units, each containing a CRA picture that can serve as an independent random access point. This segmentation allows the video stream to be divided into manageable sections that can be decoded independently, enhancing flexibility without overwhelming the decoder with entire sequence dependencies
Solution Approach 2:
CRA pictures serve as intermediary reference points between traditional IDR pictures. These intermediate access points provide decodable reference frames that mediate between random access requirements and decoding complexity, allowing flexible entry points while maintaining manageable decoder state requirements
2Adaptability or versatility
If CRA pictures with non-decodable leading pictures are allowed mid-bitstream, then the number of access points increases, but the complexity of managing leading pictures and their decodability increases
Solution Approach 1:
The decoder is prepared in advance for potential non-decodable leading pictures by implementing logic to detect and handle such cases. Flags and syntax elements are included in the bitstream to pre-indicate the presence of non-decodable leading pictures, allowing the decoder to prepare appropriate handling strategies before encountering these pictures
Solution Approach 2:
The bitstream structure itself carries information about its own decodability characteristics through flags and syntax elements. The leading pictures contain embedded metadata that enables them to self-identify their decodability status, reducing the burden on external management systems to track and manage their decodability state
3Adaptability or versatility
If mid-bitstream CRA pictures are allowed with non-decodable leading pictures, then random access flexibility is improved, but bitstream structure complexity increases
Solution Approach 1:
CRA pictures serve multiple functions simultaneously: they act as random access points, provide reference frames for subsequent decoding, and can have associated leading pictures for temporal continuity. This multi-functionality consolidates several features into a single picture type, reducing the need for multiple specialized structures
Solution Approach 2:
The bitstream structure uses parameters such as `broken_link_flag` and `rasl_pic_cnt` to dynamically indicate the presence and characteristics of non-decodable leading pictures. These parameters allow the bitstream to adapt its structure based on content requirements without fundamentally changing the underlying syntax framework
4Adaptability or versatility
If more access points are introduced for adaptive video delivery, then video delivery flexibility is improved, but the complexity of mapping units to container formats increases
Solution Approach 1:
Access units containing CRA pictures are segmented as independent mappable entities to container formats. Each access unit can be independently packaged and delivered, simplifying the mapping process compared to treating entire video sequences as single units. This segmentation aligns with container format requirements for discrete, deliverable units
Data Source
AI summary
Disclosed herein are innovations for bitstreams having clean random access (CRA) pictures and/or other types of random access point (RAP) pictures. New type definitions and strategic constraints on types of RAP pictures can simplify mapping of units of elementary video stream data to a container format. Such innovations can help improve the ability for video coding systems to more flexibly perform adaptive video delivery, production editing, commercial insertion, and the like.


