Downloads · 30 days
0
safpi-drawing-demo
taiwu2022/safpi-drawing-demo
safpi-drawing-demo is a machine learning model from taiwu2022. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as cc-by-4.0.
Interactive demo of the pipeline from "Understanding the Inner Self" (Wu et al., 2026). Upload a photo of a child's drawing and see every interpretable stage — page crop, body-part detection, facial-expression classif…
Access
Public
Updated Jul 1, 2026
Repo size
415 KB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.py453 KB · 41%
At a glance
- License
- cc-by-4.0
- Access
- Public
- Created
- Jul 1, 2026
- Updated
- Jul 1, 2026
- SHA
- 23551438
Model card
From the Hugging Face model README
Children's Drawing Feature Extractor — Demo
Interactive demo of the pipeline from "Understanding the Inner Self" (Wu et al., 2026). Upload a photo of a child's drawing and see every interpretable stage — page crop, body-part detection, facial-expression classification, colour measurement — plus the final structured feature table.
- Weights:
taiwu2022/safpi-drawing-pipeline(safpi_kaggle_combined_v1) - Code: https://github.com/UCL-Children-Drawings-Project/drawing_classification
- Runs on CPU (~10–30 s per drawing). Uploaded images are processed only for the run and not stored.
Files
| File | Purpose |
|---|---|
app.py | Gradio app; wraps run_pipeline on a single uploaded image. |
requirements.txt | CPU torch + the pipeline's third-party dependencies. |
examples/ | A few ethics-cleared sample drawings for one-click trials. |
drawing_classification/ | Vendored copy of the package (gitignored). The GitHub repo is private, so the Space cannot pip install it from git — it is shipped alongside app.py instead. Regenerate before each deploy. |
Run locally
PYTHONPATH=src python demo/app.py # uses the package from src/
Deploy as a Hugging Face Space
# 1. Vendor the package next to app.py (gitignored; refresh so it is up to date)
cp -r src/drawing_classification demo/drawing_classification
# 2. Create the Space and upload everything in demo/ (app.py, requirements.txt,
# examples/, the vendored package). Use a working proxy node if behind the GFW.
hf repos create safpi-drawing-demo --repo-type space --space-sdk gradio
caffeinate -i hf upload safpi-drawing-demo demo . --repo-type space
- License
- cc-by-4.0
- Created
- Jul 1, 2026
- Updated
- Jul 1, 2026