Recognizing 3D Actions With Shallow Convolutional Neural Networks and Quad-Joint Relative Volume Maps. IJFANS 2022, 11 (8), 3387-3391.