ishandutta0098/mukh

AI Summary: The task involves adding a new face detection model to the Mukh library. This requires finding a suitable model (small, fast, MIT/Apache licensed, PyTorch backend, high accuracy, no dlib), integrating its code into the Mukh project's structure, and ensuring proper documentation and testing. The contributor should follow the project's contribution guidelines and communicate with the maintainer before starting work.

AI Summary: Implement BlazeFace and MediaPipe landmark extraction models within the existing Mukh library framework, extending its functionality to include image and video landmark detection. This involves creating inference code based on a provided base class and adhering to the project's contribution guidelines.

AI Summary: Gather community feedback and suggestions for new features, models, pipelines, and improvements to the Mukh face analysis library. This involves reviewing suggestions, prioritizing them based on community interest and feasibility, and potentially incorporating them into the library's development roadmap.

AI Summary: The task involves modifying a Python face detection library to replace CSV output with JSON output. This includes creating a new JSON saving function, updating existing detector classes to use the new function, modifying example scripts and documentation to reflect the changes, and ensuring compatibility across three different face detection models.

AI Summary: Update the project's README and documentation to include instructions for installing the Mukh library using a Python virtual environment (venv), in addition to the existing conda instructions.