Detect "looking back" using mediapipe Python API

[monajalal]

Have I written custom code (as opposed to using a stock example script provided in MediaPipe)

None

OS Platform and Distribution

Ubuntu 22.04

MediaPipe Tasks SDK version

0.10.3

Task name (e.g. Image classification, Gesture recognition etc.)

Looking back detection

Programming Language and version (e.g. C++, Python, Java)

Python

Describe the actual behavior

Detect if the person is looking back

Describe the expected behaviour

Currently can only detect look up, down, right, left, and foreward

Standalone code/steps you may have used to try to get what you need

NA

Other info / Complete Logs

$ pip show mediapipe
Name: mediapipe
Version: 0.10.3
Summary: MediaPipe is the simplest way for researchers and developers to build world-class ML solutions and applications for mobile, edge, cloud and the web.
Home-page: https://github.com/google/mediapipe
Author: The MediaPipe Authors
Author-email: mediapipe@google.com
License: Apache 2.0
Location: /home/mona/.local/lib/python3.10/site-packages
Requires: absl-py, attrs, flatbuffers, matplotlib, numpy, opencv-contrib-python, protobuf, sounddevice
Required-by: FaceAnalyzer

How can I detect if the person is looking back if the mediapipe is running on a video camera?

I can detect look right/left, up/down, and front but I am not sure how to detect look back? I assume we can use also shoulder or arm or nose landmarks for this but I wonder if I could get some tips on this?