Research map / Detection and segmentation / CNN features
RPCANet$^{++}$: Deep Interpretable Robust PCA for Sparse Object Segmentation
Fengyi Wu, Yimian Dai, Tao Zhang, Yixuan Ding, Jian Jun Yang, Ming‐Ming Cheng, Zhenming Peng
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the Memory-Augmented Module (MAM) (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 92.62 -> 94.39 | Improved |
| replace MAM with temporary transmission (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 92.33 -> 91.52 | Hurt |
| add Deep Contrast Prior Module (DCPM) (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 90.99 -> 94.39 | Improved |
| use CBAM instead of DCPM (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 94.39 -> 91.78 | Hurt |
| increase DCPM kernel size from 9 to 17 (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 93.43 -> 94.39 | Improved |
| increase stages from 1 to 6 (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 86.69 -> 94.39 | Improved |
| increase stages from 6 to 9 (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 94.39 -> 93.97 | Hurt |
| use a single residual block instead of two (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 93.65 -> 94.39 | Improved |
| use D^(k-1) in OEM instead of only O^(k-1) (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 92.49 -> 94.39 | Improved |
| use plus operation instead of feature concatenation (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 93.27 -> 94.39 | Improved |
| add background MSE loss sigma=0.1 (ablation) | RPCANet++ · NUDT-SIRST | IoU (%): 93.78 -> 94.39 | Improved |
| add background MSE loss sigma=0.1 (ablation) | RPCANet · NUDT-SIRST | IoU (%): 88.23 -> 89.31 | Improved |
Closest papers by meaning
- Foreground Segmentation with Tree-Structured Sparse RPCA (TPAMI 2017 · similarity 0.88)
- Sparse R-CNN: An End-to-End Framework for Object Detection (TPAMI 2023 · similarity 0.85)
- Side Information in Robust Principal Component Analysis: Algorithms and Applications (ICCV 2017 · similarity 0.85)
- Sparse R-CNN: End-to-End Object Detection With Learnable Proposals (CVPR 2021 · similarity 0.85)
- Robust Kronecker Component Analysis (TPAMI 2018 · similarity 0.84)
- Projected Robust PCA with Application to Smooth Image Recovery (JMLR 2022 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
