Research map / Vision-language and multimodal / Human and embodied modeling
Expanded Parts Model for Semantic Description of Humans in Still Images
Gaurav Sharma, Frédéric Jurie, Cordelia Schmid
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 SPM with Expanded Parts Model | BoF features · Stanford 40 Actions | mAP (%): 34.9 -> 40.7 | Improved |
| replace SPM with Expanded Parts Model | BoF features · Human Attributes | mAP (%): 55.5 -> 58.7 | Improved |
| add full image context to EPM (ablation) | EPM · Willow 7 Actions | mAP (%): 66.0 -> 67.6 | Improved |
| add full image context to EPM (ablation) | EPM · Stanford 40 Actions | mAP (%): 40.7 -> 42.2 | Improved |
| add full image context to EPM (ablation) | EPM · Human Attributes | mAP (%): 58.7 -> 59.7 | Improved |
| combine EPM with AlexNet on bounding box | AlexNet (B) · Stanford 40 Actions | mAP (%): 44.0 -> 50.7 | Improved |
| combine EPM with VGG-16 on bounding box | VGG-16 (B) · Stanford 40 Actions | mAP (%): 61.3 -> 64.0 | Improved |
| combine EPM with VGG-16 on full image | VGG-16 (I) · Human Attributes | mAP (%): 55.4 -> 64.2 | Improved |
| combine EPM with VGG-16 on bounding box | VGG-16 (B) · Human Attributes | mAP (%): 64.8 -> 69.6 | Improved |
| use deep features on full image instead of bounding box (VGG-16) | VGG-16 · Stanford 40 Actions | mAP (%): 61.3 -> 70.6 | Improved |
| use deep features on full image instead of bounding box (VGG-16) | VGG-16 · Human Attributes | mAP (%): 64.8 -> 55.4 | Hurt |
| increase number of scoring parts k (ablation) | EPM · Willow 7 Actions validation | validation performance | Hurt |
Closest papers by meaning
- Attribute And-Or Grammar for Joint Parsing of Human Pose, Parts and Attributes (TPAMI 2017 · similarity 0.85)
- Instance-level Human Parsing via Part Grouping Network (ECCV 2018 · similarity 0.85)
- Learning 3D Texture-Aware Representations for Parsing Diverse Human Clothing and Body Parts (AAAI 2026 · similarity 0.85)
- "PartImageNet: A Large, High-Quality Dataset of Parts" (ECCV 2022 · similarity 0.85)
- Differentiable Multi-Granularity Human Parsing (TPAMI 2023 · similarity 0.84)
- PARTONOMY: Large Multimodal Models with Part-Level Visual Understanding (NeurIPS 2025 · similarity 0.84)
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.
