SeetaPsych Lib

A Computer Vision Toolkit for Face-based Psychological Measurement

SeetaPsych Lib is a Python library for face and body-based psychology analysis. It provides a modular Pipeline/Runner runtime and an optional Streamlit WebUI.

Requirements

• Python >= 3.10
• (Recommended) uv package manager: https://github.com/astral-sh/uv

Installation

Using uv (recommended)

uv pip install seetapsych-lib[webui] seetapsych-attributes seetapsych-configs

Using pip

Install runtime dependencies:

pip install seetapsych-lib seetapsych-attributes seetapsych-configs

If you want to run the WebUI, install extra tools manually:

pip install seetapsych-lib[webui]

Install Default Configs

# download default configs
seetapsych-manager download
# install each module requirements
seetapsych-manager setup
# download each model
seetapsych-manager cache

The setup and cache commands can be skipped. When you use the WebUI or call the library programmatically later, seetapsych-lib can install dependencies and download necessary models on demand.

Quick Start

Run WebUI (Streamlit)

seetapsych-webui --log INFO

or

python -m seetapsych_lib.webui --log INFO

A local browser window will open automatically, or you can manually navigate to: http://localhost:8501.

Common arguments:

• --dirs <DIR...>: load modules from directories
• --files <FILE...>: load modules from local config files
• --urls <URL...>: load modules from remote URLs
• --disable-builtin: disable builtin modules
• --disable-default: disable default modules
• --cache-dir <DIR>: model cache directory
• --upload-dir <DIR>: upload directory
• --log <LEVEL>: log level (e.g., DEBUG, INFO, WARNING, or an integer like 10)

Programmatic Usage

import numpy as np

from seetapsych_lib.api import Device
from seetapsych_lib.runtime import Factory, Pipeline, Runner

factory = Factory(enable_example=True)

pkg = next(p for p in factory.packages if p.name == "Example Package")

pipeline = Pipeline(factory, packages=[pkg.uid])
pipeline.add_model(pkg.uid, pkg.models[0].uid)
pipeline.solve()

ok, _ = pipeline.satisfied()
if not ok:
    pipeline.install_requirements()

runner = Runner(pipeline, device=Device("cpu"))
result = runner.run(np.zeros((64, 64, 3), dtype=np.uint8))
print(result)

Configuration

Environment Variables

The following environment variables are supported: