Refining attention weights for facial super-resolution with counterfactual attention learning
Bibliographic record
Abstract
Face Super-resolution is a challenging problem involving reconstructing High- Resolution (HR) images from Low-Resolution (LR) inputs with attention mechanisms being a widely used approach. This paper introduces counterfactual attention learning (CAL), a novel framework based on causal inference that enhances attention quality in super-resolution tasks. CAL provides a strong supervisory signal, enabling the refinement of attention mechanisms during training. Through counterfactual interventions, CAL optimizes learned attention to improve super-resolution outcomes. This method is evaluated using the Scale- Arbitrary Super-Resolution model (ArbSR), which accommodates non-integer scale factors. Experiments conducted on CelebA, FFHQ, and CMU Multi-PIE datasets across different scale factors show that CAL significantly enhances super-resolution performance. On the CMU Multi-PIE dataset, CAL improves Peak Signal-to-Noise Ratio (PSNR) by up to 13.6 % compared to baseline attention mechanisms, even under challenging variations in illumination, pose, and expression. PSNR improvement of 15.5 % was observed for CelebA dataset whereas for the FFHQ dataset, 14.5 % improvement was observed under occlusion conditions. These results highlight the robustness and effectiveness of CAL in advancing the state of super-resolution, offering substantial quantitative and qualitative improvements and showcasing its potential for face superresolution in real-world conditions.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".