Inferred Key Frames for Video Coding Session Initiation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Predictive video coding systems face significant latency and inefficiency in establishing new coding sessions due to unpredictable network bandwidth and channel errors, particularly when initial IDR frames are lost during transmission, leading to delayed video rendering and data redundancy.
Innovation Solution
The implementation of an 'inferred key frame' that is locally stored at both the encoder and decoder, allowing video coding to commence without initial frame transmission, using this frame as a prediction reference to ensure decoding and reduce latency, and enabling recovery from transmission errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If an IDR frame is transmitted to initialize a video coding session, then the decoder can decode frames properly, but transmission loss of the IDR frame causes significant latency and data redundancy
Solution Approach 1:
The patent applies preliminary action by pre-establishing shared reference frames (such as black frames or placeholder images) at both encoder and decoder before the coding session begins. This allows the decoder to immediately decode and display frames without waiting for the transmission of initial reference frames, thereby eliminating initialization latency while maintaining decoding reliability through pre-positioned fallback references.
Solution Approach 2:
The patent introduces an intermediary reference frame that acts as a mediator between the encoder and decoder during session initialization. This intermediary frame (such as a black frame or placeholder) is locally stored at both ends and serves as a temporary reference that enables decoding to proceed immediately. When actual reference frames are transmitted later, they replace the intermediary, thus bridging the initialization gap without causing latency.
2Reliability
If an IDR frame is transmitted to reset decoder state, then decoding can resume after errors, but it forces all subsequent frames to be coded predictively with respect to the IDR frame
Solution Approach 1:
The patent applies preliminary action by pre-positioning multiple reference frames (including black frames and placeholder images) at both encoder and decoder before error occurrence. When transmission errors occur, the decoder can immediately switch to using these pre-positioned frames as references without waiting for new IDR frames, maintaining error recovery capability while avoiding the need to reset coding state and lose predictive coding efficiency.
Solution Approach 2:
The patent applies discarding and recovering by allowing the system to discard corrupted reference frames and recover using pre-positioned fallback references (such as black frames or previously stored valid frames). This enables error recovery without forcing a full state reset, allowing predictive coding to continue efficiently using the recovered reference frames rather than requiring all subsequent frames to be coded predictively with respect to a new IDR frame.
Data Source
AI summary
A video coding system may initiate coding of a new coding session with reference to an “inferred key frame” that is known both to an encoder and a decoder before a coding session begins. The inferred key frame need not be transmitted between the encoder and decoder via the channel. Instead, the inferred key frame may be stored locally at the encoder and the decoder. Frames coded at the onset of a video coding session may be coded with reference to the inferred key frame, which increases the likelihood a decoder will receive a frame it can decode properly and accelerate the rate at which the decoder generates recovered video data. Inferred key frames may be used as prediction references to recover from transmission errors.