A Method to Improve Perceptual Quality of Intra- Refresh-Enabled Low-Latency Video Coding
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
A typical video encoder includes into the generated bit stream Instantaneous Decoder Refresh (IDR) units. This allows random access playback at the receiver side as well as graceful recovery from potential channel errors. Such forced IDR units typically come in repetitive patterns, which may negatively impact the perceived subjective quality if not handled properly. The reason is that the restricted encoding process of an IDR unit results in a different (regardless higher or lower) quality of reconstructed signal compared to the surrounding non-IDR ones. This causes eye-capturing irritating periodical artifacts when it occurs in patterns. This phenomenon gets to be even more pronounced when the intra refresh feature is enabled, since it forces IDR and nonIDR units to co-exist within the same picture, making the quality difference more noticeable. This paper proposes a method to hide such undesired patterns that naturally accompany the intra refresh feature. Two ideas are presented; the first one imposes restrictions that prevent unwanted fluctuations in the quantization levels between different regions of the picture, while the second hides the repetitive pattern by randomly forcing IDR blocks within specific regions of the refreshed picture. Results show that the proposed method results in improvements in subjective quality.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it