Research map / Detection and segmentation / CNN features
Liquid Fusion of Heterogeneous Representations Towards General Salient Object Detection
Ke Chen, Ling Zhou, Yi Liu, Guangqi Jiang, Gengshen Wu, Shoukun Xu
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 ConvNeXt stream to VMamba-only model (full LFNet) (ablation) | LFNet · DUTS | S-measure: 92.0 -> 93.6 | Improved |
| add VMamba stream to ConvNeXt-only model (full LFNet) (ablation) | LFNet · DUTS | S-measure: 84.8 -> 93.6 | Improved |
| add liquid fusion and SGU to naive dual-stream baseline (ablation) | LFNet · DUTS | S-measure: 92.5 -> 93.6 | Improved |
| add liquid fusion module instead of additive fusion (ablation) | LFNet · DUTS | S-measure: 92.7 -> 93.6 | Improved |
| add liquid fusion module instead of concatenation (ablation) | LFNet · DUTS | S-measure: 93.0 -> 93.6 | Improved |
| add liquid fusion module instead of cross-attention (ablation) | LFNet · DUTS | S-measure: 93.2 -> 93.6 | Improved |
| add SGU instead of bilinear upsampling (ablation) | LFNet · DUTS | S-measure: 92.9 -> 93.6 | Improved |
| add SGU instead of transposed convolution (ablation) | LFNet · DUTS | S-measure: 93.2 -> 93.6 | Improved |
| add liquid fusion module instead of additive fusion (ablation) | LFNet · VDT-2048 | S-measure: 93.5 -> 94.2 | Improved |
| add SGU instead of bilinear upsampling (ablation) | LFNet · VDT-2048 | S-measure: 93.2 -> 94.2 | Improved |
| use LFNet instead of Mamba-based Samba | Samba · DAVSOD-easy | S-measure: 81.3 -> 83.1 | Improved |
| use LFNet instead of Samba | Samba · VT5000 | S-measure: 92.8 -> 92.8 | No clear effect |
Closest papers by meaning
- Specificity-Preserving RGB-D Saliency Detection (ICCV 2021 · similarity 0.90)
- Siamese Network for RGB-D Salient Object Detection and Beyond (TPAMI 2021 · similarity 0.90)
- A Single Stream Network for Robust and Real-time RGB-D Salient Object Detection (ECCV 2020 · similarity 0.90)
- A Highly Efficient Model to Study the Semantics of Salient Object Detection (TPAMI 2021 · similarity 0.89)
- RGB-D Salient Object Detection via 3D Convolutional Neural Networks (AAAI 2021 · similarity 0.89)
- SMR-Net: Semantic-Guided Mutually Reinforcing Network for Cross-Modal Image Fusion and Salient Object Detection (AAAI 2025 · similarity 0.88)
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.
