Research map / Detection and segmentation / Weakly supervised detection and segmentation
MFNet: Multi-Filter Directive Network for Weakly Supervised Salient Object Detection
Yongri Piao, Jian Wang, Miao Zhang, Huchuan Lu
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 a directive filter to a single pseudo label Y1 (ablation) | saliency network trained on Y1 · DUTS-Test | F-measure: 0.607 -> 0.698 | Improved |
| add a directive filter to a single pseudo label Y2 (ablation) | saliency network trained on Y2 · DUTS-Test | F-measure: 0.639 -> 0.684 | Improved |
| use multiple directive filters with both labels instead of a single filter on Y1 (ablation) | MFNet · DUTS-Test | F-measure: 0.698 -> 0.710 | Improved |
| fuse labels by averaging instead of using single label Y2 (ablation) | saliency network · DUTS-Test | F-measure: 0.639 -> 0.638 | No clear effect |
| fuse labels by union instead of single label Y2 (ablation) | saliency network · DUTS-Test | F-measure: 0.639 -> 0.637 | No clear effect |
| fuse labels by intersection instead of single label Y2 (ablation) | saliency network · DUTS-Test | F-measure: 0.639 -> 0.649 | Mixed |
| use a dual-decoder with two labels instead of single label Y1 (ablation) | saliency network · DUTS-Test | F-measure: 0.607 -> 0.670 | Improved |
| replace dual decoder with multiple directive filters (ablation) | MFNet · DUTS-Test | F-measure: 0.670 -> 0.710 | Improved |
| add self-supervision between directive filters (delta 0 to 2) (ablation) | MFNet · DUTS-Test | F-measure: 0.702 -> 0.710 | Improved |
| encourage filters to learn different cues (delta -2 instead of 2) (ablation) | MFNet · DUTS-Test | F-measure: 0.710 -> 0.679 | Hurt |
| raise self-supervision weight delta from 2 to 4 (ablation) | MFNet · DUTS-Test | F-measure: 0.710 -> 0.706 | Hurt |
| saliency decoder guided by filtered cues instead of a directive filter output P1 (ablation) | MFNet · ECSSD | F-measure: 0.842 -> 0.854 | Improved |
Later papers that compared with this method
2 later accepted papers report MF-Net as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Neovascularization Segmentation via a Multilateral Interaction-Enhanced Graph Convolutional Network (TPAMI 2025 · better on 1 of 1 shared result)
- Multispectral Video Semantic Segmentation: A Benchmark Dataset and Baseline (CVPR 2023 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Weakly Supervised Salient Object Detection Using Image Labels (AAAI 2018 · similarity 0.91)
- Structure-Consistent Weakly Supervised Salient Object Detection with Local Saliency Coherence (AAAI 2021 · similarity 0.91)
- Multi-Source Weak Supervision for Saliency Detection (CVPR 2019 · similarity 0.91)
- Weakly-Supervised Salient Object Detection Using Point Supervision (AAAI 2022 · similarity 0.90)
- Learning to Detect Salient Object with Multi-source Weak Supervision (TPAMI 2021 · similarity 0.90)
- Learning to Detect Salient Objects With Image-Level Supervision (CVPR 2017 · 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.
