Research map / Detection and segmentation / CNN features
Distraction-Aware Feature Learning for Human Attribute Recognition via Coarse-to-Fine Attention Mechanism
Mingda Wu, Di Huang, Yuanfang Guo, Yunhong Wang
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 |
|---|---|---|---|
| Use Da-HAR's coarse-to-fine attention and multi-level features instead of DIAA | Human attribute recognition · WIDER-Attribute | mAP: 86.4 -> 87.3 | Improved |
| Use Da-HAR with original BCE-loss | Da-HAR · RAP | Prec: 84.54 | Improved |
| Use weighted BCE-loss to emphasize minority attributes | Da-HAR · RAP | Rec: 84.13 | Improved |
| Use weighted BCE-loss | Da-HAR · RAP | Prec: 84.54 -> 66.50 | Hurt |
| Use mixed loss to trade off Precision and Recall | Da-HAR · RAP | F1: 80.72 | Improved |
| Add multi-level feature fusion (ablation) | Da-HAR · WIDER-Attribute | mAP: 85.3 / 86.2 -> 85.7%/86.5% | Improved |
| Add the self-mask block for coarse distraction awareness (ablation) | Da-HAR · WIDER-Attribute | mAP: 85.3 / 86.2 -> 85.8%/86.6% | Improved |
| Add the masked attention block for fine distraction-aware learning (ablation) | Da-HAR · WIDER-Attribute | mAP: 85.3 / 86.2 -> 86.0%/86.8% | Improved |
| Combine multi-level feature fusion, self-mask, and masked attention (ablation) | Da-HAR · WIDER-Attribute | mAP: 85.3 / 86.2 -> 86.2%/87.1% | Improved |
| Ignore 0-label attributes during training rather than treating them as negatives (ablation) | Da-HAR · WIDER-Attribute | mAP: 86.2%/87.1% -> 86.5%/87.3% | Improved |
| Add validation data to training (ablation) | Da-HAR · WIDER-Attribute | mAP: 86.5%/87.3% -> 87.2%/88.0% | Improved |
| Add the proposed blocks to the baseline network | Da-HAR · WIDER-Attribute | inference time: 22.28ms -> 24.04ms | Hurt |
Closest papers by meaning
- Learning From Human Attention for Attribute-Assisted Visual Recognition (TPAMI 2024 · similarity 0.84)
- Hierarchical Feature Embedding for Attribute Recognition (CVPR 2020 · similarity 0.83)
- Attributes for Improved Attributes: A Multi-Task Network Utilizing Implicit and Explicit Relationships for Facial Attribute Classification (AAAI 2017 · similarity 0.83)
- Localizing by Describing: Attribute-Guided Attention Localization for Fine-Grained Recognition (AAAI 2017 · similarity 0.82)
- Deep Imbalanced Attribute Classification using Visual Attention Aggregation (ECCV 2018 · similarity 0.82)
- GUIDED: Granular Understanding via Identification, Detection, and Discrimination for Fine-Grained Open-Vocabulary Object Detection (NeurIPS 2025 · similarity 0.80)
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.
