- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
GRADE combines adaptive granularity alignment and structure-aware residual diffusion for incomplete multi-view clustering. Local correspondences and retrieved target-view prototypes guide completion. Experiments cover eight datasets and five missingness settings.
Critical strengths
- The paper is well written and easy to follow.
- Experimental settings, baseline configurations, and source-code links are clearly provided.
- Comparisons, ablations, and efficiency analyses support the method’s effectiveness.
Critical weaknesses
- Performance under high missingness needs further explanation. Since alignment depends on co-observed samples, the paper should clarify how reduced evidence affects correspondence, retrieval, and completion quality.
- Key design contributions are not fully isolated. Ablations do not directly compare adaptive versus fixed granularity or retrieved versus simpler conditioning. Key ablations and parameter analyses also use only a missing ratio of 0.9 and seed 42, limiting evidence of stability.
Questions for authors
Under higher missingness, is degradation mainly caused by weaker correspondence evidence or fewer training samples for completion? What existing observations support this explanation?
6 comments
W2:Key design contributions are not fully isolated. Ablations do not directly compare adaptive versus fixed granularity or retrieved versus simpler conditioning. Key ablations and parameter analyses also use only a missing ratio of 0.9 and seed 42, limiting evidence of stability.
We thank the reviewer for the suggestion. To further distinguish the contributions of key components, we added two ablation variants: replacing granular-ball prototypes with K-means centroids and removing the retrieved structural condition C from the diffusion model. As shown in the table, at a missing rate of 0.9 with seed 42, the full model achieves an average ACC of 53.68% across seven datasets, outperforming the two variants by 5.63 and 5.07 percentage points, respectively. These comparisons provide direct evidence of the contributions of adaptive granular-ball representations and retrieved structural conditioning.
| Variant/Dataset | Synthetic3D | EYaleB10 | COIL20MV | 100Leaves | Mfeat | LandUse-21 | Scene-15 |
|---|---|---|---|---|---|---|---|
| w/o C | 0.8266 | 0.2109 | 0.6944 | 0.3306 | 0.7285 | 0.2295 | 0.3821 |
| Kmeans C | 0.7900 | 0.2156 | 0.6153 | 0.4612 | 0.6740 | 0.2252 | 0.3822 |
| Ours | 0.8217 | 0.3000 | 0.7715 | 0.4056 | 0.8010 | 0.2443 | 0.4136 |
Furthermore, we extended the ablation study of training objectives in the main manuscript to missing rates of 0.1, 0.3, 0.5, and 0.7 across eight datasets. The results are presented in the following table. Among the 32 additional dataset–missing-rate settings, the full model achieves 19 individual best results and one tie. For example, at a missing rate of 0.7, the full model achieves an average ACC of 54.76%, outperforming the variants with only (39.09%) and only (35.19%) by 15.67 and 19.57 percentage points, respectively. These results further support the complementary roles of region alignment and diffusion-based completion across different missing rates.
The benefits of combining the two components become more pronounced as the missing rate increases. In particular, removing leads to an average ACC reduction that increases from 4.72 percentage points at a missing rate of 0.1 to 19.57 percentage points at 0.7. This indicates that region alignment becomes increasingly important for diffusion-based completion as observational evidence becomes sparser. This finding is consistent with our mechanistic analysis: Evidence-Alignment promotes cross-view semantic alignment in the shared space through region correspondence, providing a basis for the query to retrieve semantically matched local prototypes from the target view. These prototypes, in turn, provide structural guidance for Structure-Aware Residual Diffusion. Overall, the results further demonstrate the complementary roles of the two components under highly incomplete settings.
Further component ablation studies and their corresponding results are provided in our response to the following comment. All additional ablation experiments reported in this response use seed 42. Due to the time and computational budget of this revision, the remaining ablations, diffusion formulation comparisons, and parameter analyses across additional missing rates and random seeds will be included in the appendix in a subsequent revision.
Missing Rate = 0.1
| Variant/Dataset | Synthetic3D | EYaleB10 | COIL20MV | 100Leaves | Mfeat | LandUse-21 | Scene-15 | Animal |
|---|---|---|---|---|---|---|---|---|
| w/o & | 0.9366 | 0.2453 | 0.7340 | 0.7481 | 0.8280 | 0.2680 | 0.4263 | 0.2003 |
| w/o | 0.9700 | 0.4406 | 0.7590 | 0.8425 | 0.7770 | 0.2533 | 0.4644 | 0.2285 |
| w/o | 0.9616 | 0.3906 | 0.7520 | 0.7856 | 0.8070 | 0.2795 | 0.4448 | 0.2200 |
| ours | 0.9717 | 0.4437 | 0.8354 | 0.9306 | 0.8500 | 0.2748 | 0.4821 | 0.2300 |
Missing Rate = 0.3
| Variant/Dataset | Synthetic3D | EYaleB10 | COIL20MV | 100Leaves | Mfeat | LandUse-21 | Scene-15 | Animal |
|---|---|---|---|---|---|---|---|---|
| w/o & | 0.8350 | 0.1984 | 0.5416 | 0.5012 | 0.5435 | 0.2109 | 0.3215 | 0.1721 |
| w/o | 0.9316 | 0.3750 | 0.6229 | 0.6368 | 0.6330 | 0.2238 | 0.4064 | 0.1989 |
| w/o | 0.8883 | 0.3359 | 0.7097 | 0.6050 | 0.7755 | 0.2471 | 0.4057 | 0.2029 |
| ours | 0.9417 | 0.4281 | 0.8306 | 0.8860 | 0.8405 | 0.2805 | 0.4611 | 0.2312 |
Missing Rate = 0.5
| Variant/Dataset | Synthetic3D | EYaleB10 | COIL20MV | 100Leaves | Mfeat | LandUse-21 | Scene-15 | Animal |
|---|---|---|---|---|---|---|---|---|
| w/o & | 0.7950 | 0.1921 | 0.3750 | 0.3406 | 0.2605 | 0.1585 | 0.2637 | 0.1457 |
| w/o | 0.9250 | 0.3203 | 0.5340 | 0.4556 | 0.4840 | 0.2138 | 0.3451 | 0.1626 |
| w/o | 0.7233 | 0.3171 | 0.6312 | 0.4475 | 0.7080 | 0.2052 | 0.3707 | 0.1898 |
| ours | 0.9250 | 0.4250 | 0.8354 | 0.7669 | 0.8390 | 0.2733 | 0.4573 | 0.2089 |
Missing Rate = 0.7
| Variant/Dataset | Synthetic3D | EYaleB10 | COIL20MV | 100Leaves | Mfeat | LandUse-21 | Scene-15 | Animal |
|---|---|---|---|---|---|---|---|---|
| w/o & | 0.7216 | 0.1625 | 0.2993 | 0.2718 | 0.2010 | 0.1500 | 0.2113 | 0.1201 |
| w/o | 0.8350 | 0.3156 | 0.5770 | 0.3406 | 0.3600 | 0.2223 | 0.3308 | 0.1458 |
| w/o | 0.6600 | 0.2250 | 0.4625 | 0.3031 | 0.5365 | 0.1633 | 0.2954 | 0.1695 |
| ours | 0.8783 | 0.3703 | 0.8083 | 0.6194 | 0.8330 | 0.2714 | 0.4419 | 0.1579 |
We sincerely thank you for your careful review of our work. Your questions and suggestions demonstrate a deep engagement with our method and have helped us further improve the manuscript. We address each comment point by point below and describe the corresponding additional experiments and revisions.
W1:Performance under high missingness needs further explanation. Since alignment depends on co-observed samples, the paper should clarify how reduced evidence affects correspondence, retrieval, and completion quality.
We thank the reviewer for raising this concern. GRADE establishes region correspondences based on co-observed samples and uses them to learn cross-view alignment in the shared space. It then constructs diffusion conditions by matching the query with target-view prototypes. Therefore, as the number of co-observed samples decreases, the statistical evidence for region correspondence becomes limited. The resulting uncertainty may propagate to shared-space alignment, affecting the semantic matching between queries and retrieved prototypes and weakening the structural guidance provided to the diffusion model.
Meanwhile, fewer samples jointly observed in the target view and other views also limit the training information available to the diffusion model. Thus, performance degradation under high missing rates may involve both the reliability of correspondences and retrieval conditions and the sufficiency of diffusion training samples. To further distinguish their effects, we report controlled experiments in Q1 that separately reduce region correspondence evidence, diffusion training samples, and both simultaneously, and analyze their respective impacts on performance.