Skip to main content

Paper - Pytorch

Project description

ClipQ (WIP)

An easy-to-use interface for experimenting with OpenAI's CLIP model by encoding image quadrants. By splitting images into quadrants and encoding each with CLIP, we can explore how the model perceives various parts of an image.


Table of Contents


Install the package via pip:

pip install clipq


Here's a brief example to get you started:

from clipq.main import CLIPQ

test = CLIPQ(query_text="A photo of a cat")

#input, url => embed
vectors = test.run_from_url(url="", h_splits=3, v_splits=3)




  1. Fork the repository on GitHub.
  2. Clone the forked repository to your machine.
  3. Create a new branch with an appropriate name.
  4. Make your changes and commit with a meaningful commit message.
  5. Push your changes to your forked repository.
  6. Create a Pull Request against the original repository.


This project is licensed under the MIT License - see the LICENSE file for details.


  • Output captions of all 4 quadrants
  • Make captions using any of the following: openclip G, OpenCLIP G or siglip L or EVA G
  • Image Division: Ability to split an image into quadrants (2x2). Extended ability to split an image into 9 equal parts (3x3).
  • Vector Representation: Generation of a CLIP vector for the entire image and individual CLIP vectors for each split part or quadrant.
  • Sub-clip Concerns: Identification of hard chunking issues with standard quadrant splitting.
  • Noise Reduction: Introduction of non-standard shapes (possibly polygons) for image parts to reduce noise. Aim to tackle interlacing issues during upscaling.
  • Upscaling: Address potential tiling issues during the upscaling process.
  • Flexibility in Sub-clipping: Configurable options to choose between 2x2 or 3x3 image division.
  • Prior Training: Training mechanism to use the data of quadrant CLIP vectors.

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

clipq-0.0.7.tar.gz (5.2 kB view hashes)

Uploaded Source

Built Distribution

clipq-0.0.7-py3-none-any.whl (5.2 kB view hashes)

Uploaded Python 3

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