The library for solving frequently encountered problems in the field of computer vision

With pip

pip3 install git+https://github.com/VasilyBoychuk/cvints.git

The library has several main entities:

• Dataset
• Model
• Processing results
• Metrics
• Experiment

You can store information about train/validation/test dataset in the Dataset-like classes. To do this, specify the paths to images and annotation files.

You can define model using model meta-data like input images size, post-processing methods, etc

Store model processing results with useful images info to chose post-processing methods to complete and evaluate your model

Chose some existed metrics for you Task, or implement some specific one.

You can think of the entire process of evaluating the model's performance with data as an experiment in which all actions with other entities take place

• Check dataset before start testing model with it (rus | eng)
• Calculate metrics and check model results on an object detection task (rus | eng)