- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This article introduces SOLA-SR, which organizes feature channels into nested subspaces through learnable orthogonal transformations, and dynamically selects some channels based on the super-resolution scale to reduce decoding computations at lower scales. The method is combined with a LIIF-style decoder and implements selective decoding through algebraic transformations. Experiments on several standard datasets and different upscaling ratios have verified both the reconstruction quality and running efficiency, and the numerical consistency between dense and selective decoding is also reported.
Critical strengths
- The method design is simple and theoretically solid. Selective decoding is based on precise algebraic transformations rather than approximate calculations, and experiments have confirmed its consistency with dense decoding.
- The scale-related feature organization is somewhat novel. A shared image encoding can support different scales while reducing the decoder's computation at lower magnifications.
Critical weaknesses
- The core ablation doesn't fully prove the role of the learned orthogonal basis. In Table 4, the learned basis is optimized separately after freezing the encoder/decoder, while the identity/random baseline doesn’t undergo fully matched joint training, so the performance boost can’t be clearly attributed to the orthogonal basis itself.
- Efficiency advantages are limited and there’s a lack of thorough comparison with recent methods. At ×4 and higher magnifications, the gate is almost fully open and the computational gain is minimal; moreover, there’s no fair comparison with recent efficient ASSR methods. Therefore, it’s not yet sufficient to demonstrate that the efficiency advantage is competitive.
Questions for authors
Is it possible to provide a completely fair ablation: the identity, random orthogonal, and learned orthogonal settings all use the same encoder/decoder, loss, training budget, and scale distribution for joint training?
Could you also add a fair efficiency comparison with recent efficient ASSR methods, and clearly report the full training setup and parameter count?
In the appendix, the number of different scales and the number of experiment pairs seem inconsistent; could you clarify the evaluation protocol?
4 comments
B. Recent methods establish the quality and efficiency comparison.
We evaluated quality on the same images at ×2, ×3, and ×4, and timed across 200 scales on the same experiment settings. In accordance with your requirements, we have incorporated the recent efficient ASSR method as a comparison.
| Method | Parameters (M) | Avg. Y-PSNR ↑ | Avg. LPIPS ↓ | Independent (s) ↓ | Encoding reuse (s) ↓ |
|---|---|---|---|---|---|
| SOLA-SR Exact | 22.323 | 29.686 | 0.1319 | 18.526 | 10.803 |
| AnySR–EDSR | 2.297 | 29.301 | 0.2004 | 12.681 | — |
| DIIF–RDN small | 22.305 | 28.979 | 0.2224 | 10.723 | — |
| HIIF–RDN | 23.709 | 29.137 | 0.1891 | 34.802 | 27.278 |
| GSASR–RDN | 41.198 | 29.157 | 0.1878 | 98.161 | 89.682 |
| Thera RDN-Air | 21.982 | 29.418 | 0.2022 | 18.273 | 9.590 |
These results show that SOLA-SR reduces inference time through selective decoding while delivering competitive performance among the evaluated ASSR methods.
The added comparison with recent ASSR methods and the accompanying discussion appear in the dark-orange text with a dark-blue underline around Table 4 (p. 8, lines 410–425).
For Questions:
1.Could you provide a fully fair ablation in which the identity transform, fixed random orthogonal transform, and learned orthogonal transform are all jointly trained using the same encoder, decoder, losses, training budget, and scale distribution?
Our reply:
OK,we trained the identity (T5), fixed random orthogonal (T6), and learned orthogonal (T4) variants with the same encoder and decoder architecture, initialization, gate, data, loss, scale distribution, and training budget.
| Dataset | Scale | T4 − T5 ΔY-PSNR (dB) | T4 − T6 ΔY-PSNR (dB) | T4 − T5 ΔLPIPS | T4 − T6 ΔLPIPS |
|---|---|---|---|---|---|
| Set5 | ×2 | +0.194 ± 0.006 | +0.177 ± 0.007 | −0.00024 ± 0.00005 | −0.00041 ± 0.00005 |
| Set5 | ×3 | +0.073 ± 0.014 | +0.052 ± 0.005 | +0.00085 ± 0.00076 | −0.00367 ± 0.00029 |
| Set5 | ×4 | +0.003 ± 0.002 | +0.022 ± 0.009 | −0.00265 ± 0.00033 | −0.00443 ± 0.00043 |
| Set14 | ×2 | +0.137 ± 0.018 | +0.111 ± 0.025 | −0.00048 ± 0.00028 | +0.00018 ± 0.00038 |
| Set14 | ×3 | +0.041 ± 0.002 | +0.022 ± 0.003 | +0.00176 ± 0.00081 | −0.00380 ± 0.00058 |
| Set14 | ×4 | +0.015 ± 0.002 | +0.016 ± 0.002 | −0.00240 ± 0.00046 | −0.00642 ± 0.00025 |
| B100 | ×2 | +0.080 ± 0.003 | +0.066 ± 0.001 | −0.00103 ± 0.00035 | +0.00008 ± 0.00039 |
| B100 | ×3 | +0.028 ± 0.002 | +0.013 ± 0.002 | −0.00123 ± 0.00100 | −0.00486 ± 0.00055 |
| B100 | ×4 | +0.004 ± 0.003 | +0.013 ± 0.001 | −0.00369 ± 0.00048 | −0.00421 ± 0.00034 |
| Urban100 | ×2 | +0.245 ± 0.008 | +0.182 ± 0.006 | −0.00176 ± 0.00006 | −0.00129 ± 0.00012 |
| Urban100 | ×3 | +0.064 ± 0.005 | +0.054 ± 0.004 | −0.00064 ± 0.00041 | −0.00142 ± 0.00020 |
| Urban100 | ×4 | +0.037 ± 0.003 | +0.038 ± 0.002 | −0.00235 ± 0.00030 | −0.00203 ± 0.00037 |
After matched joint training, the learned basis showed better overall performance than both fixed bases across dataset–scale comparisons.
2.Could you add a fair efficiency comparison with recent efficient ASSR methods and explicitly report the full training settings and parameter counts?
Our reply:
A. The efficiency comparison covers recent ASSR methods.
| Method | Parameters (M) | Avg. Y-PSNR ↑ | Avg. LPIPS ↓ | Independent (s) ↓ | Encoding reuse (s) ↓ |
|---|---|---|---|---|---|
| SOLA-SR Exact | 22.323 | 29.686 | 0.1319 | 18.526 | 10.803 |
| AnySR–EDSR | 2.297 | 29.301 | 0.2004 | 12.681 | — |
| DIIF–RDN small | 22.305 | 28.979 | 0.2224 | 10.723 | — |
| LM-LIIF–EDSR-b | 1.386 | 29.460 | 0.2073 | 3.003 | 2.578 |
| HIIF–RDN | 23.709 | 29.137 | 0.1891 | 34.802 | 27.278 |
| GSASR–RDN | 41.198 | 29.157 | 0.1878 | 98.161 | 89.682 |
| Thera RDN-Air | 21.982 | 29.418 | 0.2022 | 18.273 | 9.590 |
Compared with the same-weight Dense path, Exact reduces sequence time by 6.1% with independent encoding and 11.1% with encoding reuse while preserving reconstruction quality. The broader comparison places SOLA-SR’s measured quality and efficiency alongside recent methods.
B.The training setup settings for both models.
| Setting | Matched protocol |
|---|---|
| Training data | 780 DIV2K training images |
| Input and queries | 48×48 LR input, 192×192 HR crop, 64×64 query region |
| Batch and scales | Batch 4; scales sampled from , with ×4 every 20 steps |
| Training budget | 20k decoder-only warm-up steps, then 80k joint steps |
| Optimizer | AdamW, weight decay 0, gradient clipping at 1 |
| Learning rates | Warm-up ; joint stage for encoder and decoder and for SOLA-SR’s Cayley parameters, with stage-wise cosine decay |
| Loss weights | Charbonnier 1, LPIPS 0.02, color 0.01; scheduled Sobel term |
3.The number of scales and the number of experimental pairs reported in the appendix appear inconsistent. Could you clarify the evaluation protocol?
Our reply:
We checked the experiment server and confirmed that the appendix retained an outdated total count of experimental pairs. We confirm that all experiments compared within each table used the same evaluation protocol. To clarify the counts, each scale uses 219 pairs, giving a total of 657 image–scale pairs per method across the fixed scales ×2, ×3, and ×4. Appendix Table 17 still reported the earlier total of 876 pairs across four scales. This was a counting error, which we have now corrected. Thank you for pointing it out.
For the evaluation pair count, please consult the plain black text in §4.1 (p. 6, lines 275–280), which reports 657 integer-scale image–scale pairs.
Thank you for reading our paper carefully and pointing out where it could be improved. Below, we address the concerns you raised and respond to each of your questions.
We have also revised the manuscript. Please look for \textcolor[RGB]{166,75,0}{\text{dark-orange text}} or \textcolor[RGB]{166,75,0}{\underline{\text{dark-orange underlining}}} to identify revisions addressing your comments.
For Critical Weaknesses:
1. The core ablation does not sufficiently establish the role of learning the orthogonal basis. In Table 4, the learned basis is optimized separately with the encoder and decoder frozen, while the identity and random-basis controls do not undergo fully matched joint training. The performance gains therefore cannot be clearly attributed to the orthogonal basis itself.
Our response:
Thank you for pointing out this limitation of Table 4. We jointly trained the encoder and decoder with an identity basis (T5), a fixed random orthogonal basis (T6), or a learned orthogonal basis (T4). The three variants used the same initialization, gate, data, loss, scale distribution, and training budget across three matched seeds.
The learned basis achieved comparative performance than both fixed bases in all 12 dataset–scale comparisons after matched joint training, supporting its contribution within the gated model.
You can find the clarification of which components are jointly trained in the dark-red text with a dark-blue underline in §3.1 (p. 3, lines 154–157). The separate frozen-backbone study is described in the plain black text in §4.3 (p. 8, lines 430–431).
2. The efficiency advantage is limited, and comparisons with recent methods are insufficient. At ×4 and higher scales, the gate is almost fully open, leaving little computational benefit. Fair comparisons with recent efficient ASSR methods are also missing. The evidence is therefore insufficient to establish a competitive efficiency advantage.
Our response:
We address the measured efficiency gain and the comparison with recent ASSR methods separately.A and B.
A. Selective decoding yields its largest savings at lower magnifications.
We compared Dense and Exact with identical weights over 200 scales, reporting the results below
Above ×3.5, all channels are active and the first-layer width saving is zero. The remaining measured time difference reflects the complete Dense and Exact execution paths. The larger reductions at lower scales are the benefit targeted by selective decoding.