Research map / Detection and segmentation / CNN features
On Symbiosis of Attribute Prediction and Semantic Segmentation
Mahdi M. Kalayeh, Mubarak A. Shah
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 semantic segmentation-based pooling (SSP) instead of average pooling | VGG16-like attribute network · CelebA (original) | classification error: 9.83 -> 8.98 | Improved |
| add semantic segmentation-based gating (SSG) instead of average pooling | VGG16-like attribute network · CelebA (original) | classification error: 9.83 -> 9.13 | Improved |
| combine SSP and SSG predictions (ablation) | SSP · CelebA (pre-cropped) | AP: 81.02 -> 81.45 | Improved |
| use semantic masks instead of landmark bounding boxes in SSP (ablation) | SSP · CelebA (pre-cropped) | AP: 79.95 -> 81.02 | Improved |
| pool with spatial pyramid pooling instead of semantic regions (ablation) | average pooling baseline · CelebA (pre-cropped) | AP: 79.74 -> 77.69 | Hurt |
| feed segmentation maps as extra input channels (ablation) | average pooling baseline · CelebA (pre-cropped) | AP: 79.74 -> 79.74 | No clear effect |
| add SSP + SSG instead of average pooling | VGG16-like attribute network · LFWA | AP: 82.69 -> 85.28 | Improved |
| add Symbiotic Augmentation to Inception-V3 | Inception-V3 · CelebA (original) | AP: 79.28 -> 80.10 | Improved |
| add Symbiotic Augmentation to Inception-V3 | Inception-V3 · WIDER Attribute | AP: 85.86 -> 87.58 | Improved |
| add Symbiotic Augmentation to Inception-V3 | Inception-V3 · Berkeley Attributes of People | AP: 92.87 -> 94.80 | Improved |
| add Symbiotic Augmentation to Inception-V3 | Inception-V3 · LFWA | AP | No clear effect |
| initialize face parsing from attribute-pretrained weights instead of ImageNet | Inception-V3 face parsing · Helen Face | parsing performance | Improved |
Closest papers by meaning
- Improving Facial Attribute Prediction Using Semantic Segmentation (CVPR 2017 · similarity 0.94)
- Automatic Discovery, Association Estimation and Learning of Semantic Attributes for a Thousand Categories (CVPR 2017 · similarity 0.89)
- Attributes for Improved Attributes: A Multi-Task Network Utilizing Implicit and Explicit Relationships for Facial Attribute Classification (AAAI 2017 · similarity 0.88)
- Learning Spatial-Semantic Relationship for Facial Attribute Recognition With Limited Labeled Data (CVPR 2021 · similarity 0.87)
- Learning Attributes Equals Multi-Source Domain Generalization (CVPR 2016 · similarity 0.86)
- SegFace: Face Segmentation of Long-Tail Classes (AAAI 2025 · similarity 0.85)
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.
