Research map / Image generation / Image restoration
Image Saliency Prediction in Transformed Domain: A Deep Complex Neural Network Method
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use Sal-DCNN instead of SAM | saliency prediction · CAT2000 | KL divergence: 1.12 -> 0.63 | Helped |
| use Sal-DCNN instead of SAM | saliency prediction · MIT1003 | CC: 0.61 -> 0.62; NSS 2.19 -> 2.10 | Mixed |
| add transformed domain (complex encoder) versus real-valued DenseNet (ablation) | Sal-DenseNet · CAT2000 | KL divergence: 0.97 -> 0.74 | Helped |
| add phase decoder stream (ablation) | Sal-DCNN-P · CAT2000 | KL divergence: 0.74 -> 0.65 | Helped |
| add residual amplitude stream (ablation) | Sal-DCNN-PP · CAT2000 | KL divergence: 0.65 -> 0.63 | Helped |
| add squared-leakage ReLU (ablation) | Sal-DCNN · CAT2000 | KL divergence: 0.01 reduction | Helped |
| add channel-wise polynomial projection (ablation) | Sal-DCNN · CAT2000 | KL divergence: 0.03 reduction | Helped |
| replace amplitude spectrum with fixed value or white Gaussian noise (ablation) | fixation map recovery from phase · MIT1003 | CC: 0.94 -> 0.41 (fixed), 0.01 (WGN) | Hurt |
| add white Gaussian noise of 0.3 sigma to phase spectrum (ablation) | fixation map recovery from phase · MIT1003 | CC: CC = 0.94 | No clear effect |
Closest papers by meaning
- Shallow and Deep Convolutional Networks for Saliency Prediction (CVPR 2016 · similarity 0.90)
- Non-Local Deep Features for Salient Object Detection (CVPR 2017 · similarity 0.88)
- Understanding and Visualizing Deep Visual Saliency Models (CVPR 2019 · similarity 0.87)
- Energy-Based Generative Cooperative Saliency Prediction (AAAI 2022 · similarity 0.87)
- Recurrent Attentional Networks for Saliency Detection (CVPR 2016 · similarity 0.87)
- Mesh Saliency: An Independent Perceptual Measure or a Derivative of Image Saliency? (CVPR 2021 · similarity 0.87)
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.