Research map / Detection and segmentation / CNN features
Rapid Salient Object Detection With Difference Convolutional Neural Networks
Zhuo Su, Li Liu, Matthias Müller, Jiehua Zhang, Diana Wofk, Ming‐Ming Cheng, Matti Pietikäinen
Read the paperCheck your idea against 137,217 papers
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 STDM temporal module (sc+cd+ad) (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .705 -> .755 | Improved |
| use CSTDC instead of standard convolution in STDM (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .724 -> .739 | Improved |
| use ASTDC instead of standard convolution in STDM (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .724 -> .747 | Improved |
| increase frames from 2 to 8 (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .721 -> .755 | Improved |
| increase frames from 4 to 8 (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .751 -> .755 | No clear effect |
| use replicate padding instead of zero padding (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .731 -> .755 | Improved |
| use STDM instead of temporal attention (ablation) | STDNet-A · VSOD DAVSOD | S-measure: .732 -> .755 | Improved |
| swap in EfficientNet-B5 backbone | STDNet-A · VSOD DAVSOD | S-measure: .755 -> .768 | Improved |
| use SDNet difference convolutions instead of standard baseline (scale x0.75) (ablation) | SDNet · ISOD ECSSD (scale x0.75) | F-measure: .887 -> .897 | Improved |
| use SDNet instead of PiDiNet difference convolution (scale x1.0, 240 input) (ablation) | SDNet · ISOD ECSSD | F-measure: .877 -> .890 | Improved |
| use SDNet instead of baseline at scale x1.0, 320 input (ablation) | SDNet · ISOD ECSSD | F-measure: .904 -> .903 | No clear effect |
| use pre-trained ImageNet backbone with SDNet-A | SDNet-A · ISOD ECSSD | S-measure: .893 -> .908 | Improved |
Closest papers by meaning
- Salient Object Detection in the Deep Learning Era: An In-Depth Survey (TPAMI 2021 · similarity 0.91)
- Deep Contrast Learning for Salient Object Detection (CVPR 2016 · similarity 0.90)
- DVSOD: RGB-D Video Salient Object Detection (NeurIPS 2023 · similarity 0.89)
- Recurrently Aggregating Deep Features for Salient Object Detection (AAAI 2018 · similarity 0.89)
- Deep Embedding Features for Salient Object Detection (AAAI 2019 · similarity 0.89)
- Disentangled High Quality Salient Object Detection (ICCV 2021 · similarity 0.89)
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.
