Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model
Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, Artsiom Sanakoyeu
arXiv
This research note highlights the use of raw HSV (Hue, Saturation, Value) colorspace representation (capturing chromatic-luminance), and true-texture (matrix-based) representation of images for image mining applications in the social sciences. Specifically: we focus here on the basics of teaching computers to ‘think like people’ in making decisions about what visual content is most interesting or important to human viewers. Our examples capture the facts that, (a) computers see ‘colors as numbers’, rather than as meaningful sections of an image, and (b) computers see texture as ‘numbers’, rather than as meaningful ‘hard’ or ‘soft’ sections of an image. Illustrations are provided using the R packages ‘colorfindr’, ‘glcm’, ‘imager’ and ‘plotly’.
Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, Artsiom Sanakoyeu
Bilge Acun, Benjamin Lee, Fiodar Kazhamiaka, Kiwan Maeng, Manoj Chakkaravarthy, Udit Gupta, David Brooks, Carole-Jean Wu
Ilkan Esiyok, Pascal Berrang, Katriel Cohn-Gordon, Robert Künnemann