Research map / Detection and segmentation / CNN features
Predicting When Saliency Maps Are Accurate and Eye Fixations Consistent
Check 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 image features from an ImageNet-trained AlexNet to predict saliency-map accuracy | Saliency-map algorithms on MIT and PASCAL · MIT and PASCAL | Spearman correlation with true saliency accuracy: 0.52 on MIT; 0.80 on PASCAL | Helped |
| Use predicted per-image accuracy to select the saliency algorithm | Saliency algorithm selection · MIT | saliency prediction performance: about 1% | Helped |
| Predict fixation consistency from image features based on object recognition | Eye-fixation consistency prediction · MIT | Spearman correlation with true consistency | Helped |
| Use saliency-map accuracy as a proxy for fixation consistency | Saliency models · MIT | Spearman correlation with fixation consistency: around 0.2 − 0.3; around 0.82 | Mixed |
| Use fixation-map entropy as a measure of fixation consistency | Eye-fixation consistency measurement · MIT | correlation with consistency measure: about 0.85 | Mixed |
| Use saliency-map entropy to infer fixation consistency | Saliency models · MIT | correlation with fixation consistency: ≤0.25 | No clear effect |
Closest papers by meaning
- Saliency Unified: A Deep Architecture for Simultaneous Eye Fixation Prediction and Salient Object Segmentation (CVPR 2016 · similarity 0.92)
- Understanding and Visualizing Deep Visual Saliency Models (CVPR 2019 · similarity 0.90)
- A Deeper Look at Saliency: Feature Contrast, Semantics, and Beyond (CVPR 2016 · similarity 0.90)
- Salient Object Detection Driven by Fixation Prediction (CVPR 2018 · similarity 0.90)
- An Integrated Model for Effective Saliency Prediction (AAAI 2017 · similarity 0.89)
- What Do Deep Saliency Models Learn about Visual Attention? (NeurIPS 2023 · 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.