Adaptive unequal error protection for subband image 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
An adaptive subband image coding system is proposed to investigate the performance offered by implementing unequal error protection among the subbands and within the subbands. The proposed system uses DPCM and PCM codecs for source encoding the individual subbands, and a family of variable rate channel codes for forward error correction. A low resolution family of trellis coded modulation codes and a high resolution family of punctured convolutional codes are considered. Under the constraints of a fixed information rate, and a fixed transmission bandwidth, for any given image, the proposed system adaptively selects the best combination of channel source coding rates according to the current channel condition. Simulations are performed on the AWGN channel, and comparisons are made with corresponding systems where the source coder is optimized for a noiseless transmission (classical optimization) and a single channel code is selected. Our proposed joint source-channel systems greatly outperform any of the nonadaptive conventional nonjoint systems that use only a single channel code at all channel SNRs, extending the useful channel SNR range by an amount that depends on the code family. A nonjoint adaptive equal error protection system is considered which uses the classically optimized source codec, but chooses the best single channel code for the whole transmission according to the channel SNR. Our systems outperform the corresponding adaptive equal error protection system by at most 2 dB in PSNR; and more importantly, show a greater robustness to channel mismatch. It is found that most of the performance gain of the proposed systems is obtained from implementation of unequal error protection among the subbands, with at most 0.7 dB in PSNR additional gain achieved by also applying unequal error protection within the subbands. We use and improve a known modeling technique which enables the system to configure itself optimally for the transmission of an arbitrary image, by only measuring the mean of lowest frequency subband and variances of all the subbands.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| 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