Skip to main content

Python library for various computer vision problems with a focus on easy usage

Project description

Build status


ceevee (read like CV, i.e. computer vision) is a Python library for various computer vision problems with a focus on easy usage.

ceevee aims to be a bridge between deep learning practitioners training accurate models and product-oriented software engineers who just want to process their images instead of diving into the deep learning ecosystem.

Python 3.6+ is supported.


From PyPI - not available yet

From source

python bdist_wheel
pip install -U ceevee-0.0.1-py3-none-any.whl



All tasks shares the same API


python -m ceevee.cli task /path/to/img1.jpg /path/to/img2.jpg ... /path/to/imgN.jpg > result.json


HTTP API is based on Falcon, so it can be used with any WSGI server, such as uWSGI or Gunicorn.

  • install your favourite WSGI server (e.g. pip install gunicorn)
  • set env variable CEEVEE_TASKS for your tasks, multiple comma separated tasks are supported, e.g. CEEVEE_TASKS=task1,task2
  • run a server CEEVEE_TASKS=dummy gunicorn ceevee.cv_http;
  • send a POST request with image parameter.
$ http -f POST localhost:8000/dummy image@/tmp/img.jpg
HTTP/1.1 200 OK
Connection: close
Date: Sat, 14 Sep 2019 13:47:39 GMT
Server: gunicorn/19.9.0
content-length: 37
content-type: application/json

    "result": [
    "success": true

Python API

from ceevee.utils import read_img
from ceevee.dummy import DummyPredictor
baseline = DummyPredictor()
img = read_img('/path/to/img.jpg')
result = baseline(img)


Yes, you can add a new model!


  • create a GitHub issue with your suggested model;
  • create a new Baseline class (see ceevee.dummy.DummyBaseline) and implement three methods (preprocess, process, postprocess);
  • add your model to MODELS at ceevee/
  • add tests to tests/;
  • once CI is green, create a pull request!


  • infrastructure:
    • packaging, pip
  • APIs:
    • http: tests, error handling
  • models:
    • face detection
    • face emotion
    • face keypoints
    • car detection
    • crowd density estimation

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ceevee-0.0.1.tar.gz (5.0 kB view hashes)

Uploaded source

Built Distribution

ceevee-0.0.1-py2.py3-none-any.whl (5.7 kB view hashes)

Uploaded py2 py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page