Digital Video Signal Modification TRIZ Cases

Eye-gaze-driven foveated compression preserves image quality at the focal region while cutting AR memory and processing load.

Spatial neighbor and motion prediction cues narrow inter prediction contexts, cutting motion vector bits while preserving decoding accuracy.

Unified offset mapping lets MIP prediction handle different block sizes with less storage and processing time while preserving coding efficiency.

Predefined intra block copy filters and selective transforms improve video compression while avoiding unstable parameter tuning and excess compute.

Adaptive mapping of residual samples lowers entropy in video coding, improving compression efficiency and reconstructed video quality.

Motion-guided temporal chroma filtering derives offsets from reference-picture samples to reduce artifacts and improve coding efficiency.

Distribution functions decode out-of-range bitstream values to preserve perceptual quality while keeping image and video compression efficient.

Refined motion information reorders DMVR candidate lists and removes redundant entries to improve video coding efficiency.

Inter-layer reference flags help layered video decoders choose the right pictures, improving coding efficiency while reducing processor, memory, and network use.

Sequential one-sample-wide block partitions improve intra prediction accuracy while keeping codec signaling overhead low.

Adaptive filter selection improves inter-image prediction, cutting residual energy and bitrate while preserving reconstructed image quality.

Wrap-around block vector difference coding improves intra block copy prediction to cut video bitrate while preserving reconstruction quality.

Pre-filling HMVP tables with default motion candidates improves inter prediction efficiency while lowering video coding bandwidth and runtime complexity.

Dynamic Rice parameter derivation from context-coded bins improves transform coefficient coding efficiency while reducing residual complexity and memory use.

Cross-component model-based taps improve chroma adaptive loop filtering by deriving non-integer sample inputs to reduce artifacts and complexity.

Subblock affine seed vectors improve motion prediction in video decoding, raising compression efficiency beyond HEVC limits.

Combining subblocks with similar motion vectors lets BDOF run once per merged region, improving coding efficiency with lower processing load.

Pre-trained in-loop filter parameters tailored to videoconference content improve visual quality and coding efficiency without raising bit rate.

Transform coefficient statistics limit MTS candidates by block, reducing transform complexity while improving image encoding efficiency.

Template-matching block vector refinement improves intra block copy prediction accuracy and video compression with adaptive motion vector precision.

Frame type, sub-block type, and scene-aware offset skipping cut sample adaptive offset overhead while preserving coding precision on mobile devices.

Temporal scaling adapts a guiding motion vector from an unavailable reference picture, improving block prediction accuracy without extra signaling.

Integer correction bounds keep filtered transform coefficients within quantization limits, improving video reconstruction MSE and fidelity.

Adaptive prediction granularity improves coding block prediction accuracy, lowers residual values, and supports better subjective lossless reconstruction.

Predetermined MTS indices let the decoder infer LFNST use, cutting transform tests and signaling bits with minimal coding loss.

Recalculated intra prediction modes from reconstructed blocks improve future MPM and OBIC coding while limiting signaling overhead.

Adaptive thresholds from reference-sample changes let decoders switch prediction coefficients and improve edge-aware image prediction efficiency.

Low-rank matrix factorization cuts MIP storage and computation in video coding while preserving block prediction accuracy.

Multiple neighboring quantization parameters predict local QP values, cutting signaling overhead while preserving consistent visual quality.

Shared memory lets video decoding cores track completed samples before intra prediction, reducing waits and improving parallel coding efficiency.

Separate luma and chroma partitioning cuts coding overhead and improves video compression by better using chroma-luma correlation.

Constrained LMCS codewords and simplified luma-chroma scaling improve dual-tree video coding efficiency while reducing complexity and latency.

Boundary downsampling and selective upsampling improve matrix-based intra prediction, balancing video compression gains with processing complexity.

Adaptive sub-block coding order improves intra prediction by lowering bit rate while preserving video quality under bandwidth and memory limits.

Weighted blending of filtered margins and decoder-side mode selection improves overlapping pixel quality without full neural filtering cost.

Weighted template partitioning improves video block prediction accuracy while limiting matching complexity to boost compression efficiency.

Separate quantization matrices for different block sizes retain frequency-based coefficient control and improve subjective image quality.

A unified intra prediction mode list improves image compression efficiency while lowering hardware and software complexity in decoding.

Screening out unimportant and exceptional reference samples cuts video coding complexity and bandwidth while improving color prediction accuracy.

Ordered neighboring-block availability checks enable parallel merge candidate list setup, improving motion compensation efficiency in video decoding.

Conditional SPS flags include PTL, DPB, and HRD data only when needed, reducing bitstream overhead while preserving decoding accuracy.

Preselected subpicture segments are recombined into access units for low-delay streaming, smooth view switching, and personalized playback.

Multiple neighboring blocks guide CABAC context selection, improving compression efficiency and reconstructed video block quality.

BDOF refines video prediction when one motion predictor falls outside the reference picture, cutting residuals and bitrate.

A decoder uses quantization coefficient statistics to infer transform skip without CU flags, reducing bit overhead in video coding.

A two-stage transform with quad-splits improves intra subpartition coding by condensing non-zero coefficients and lowering bit rates.

Separate CABAC probability models for subblock merge and inter affine modes improve motion vector coding efficiency while limiting decoding complexity.

Band-offset-only chroma filtering uses quantized luma samples to cut reconstruction errors and preserve video quality in lossy coding.

Different quantization matrices for each block size preserve frequency-based coefficient control and improve decoded image quality.

Block-specific inverse quantization preserves frequency-based control after coefficient zeroing, improving decoded image quality.