Research map / Vision-language and multimodal / Human and embodied modeling
Attention-Aware Polarity Sensitive Embedding for Affective Image Retrieval
Xingxu Yao, Dongyu She, Sicheng Zhao, Jie Liang, Yu-Kun Lai, Jufeng Yang
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 |
|---|---|---|---|
| replace N-pair loss with polarity-sensitive EP loss (ablation) | ResNet-50 embedding · FI | mAP8: 0.5217 -> 0.5680 | Improved |
| replace N-pair loss with EP loss (ablation) | ResNet-50 embedding · FI | mAP2: 0.8062 -> 0.8558 | Improved |
| add attention loss (AT) to EP loss (ablation) | EP loss · FI | mAP8: 0.5680 -> 0.6225 | Improved |
| add attention loss (AT) to EP loss (ablation) | EP loss · FI | mAP2: 0.8558 -> 0.7816 | Hurt |
| replace EP loss with attention-weighted WEP loss in the full model (ablation) | APSE · FI | mAP8: 0.7190 -> 0.7344 | Improved |
| replace EP loss with WEP loss in the full model (ablation) | APSE · FI | mAP2: 0.8912 -> 0.9079 | Improved |
| embedding losses instead of softmax fine-tuning | ResNet · FI | mAP8: 0.4380 -> 0.5217 | Improved |
| set trade-off lambda between WEP and attention loss to 0.5 (ablation) | APSE · FI | mAP8 and mAP2 | Improved |
Closest papers by meaning
- Retrieving and Classifying Affective Images via Deep Metric Learning (AAAI 2018 · similarity 0.89)
- Emotional Attention: A Study of Image Sentiment and Visual Attention (CVPR 2018 · similarity 0.87)
- EmoEdit: Evoking Emotions through Image Manipulation (CVPR 2025 · similarity 0.83)
- Multiple Feature Refining Network for Visual Emotion Distribution Learning (AAAI 2025 · similarity 0.83)
- MART: Masked Affective RepresenTation Learning via Masked Temporal Distribution Distillation (CVPR 2024 · similarity 0.83)
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models (ICML 2025 · similarity 0.83)
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.
