Researchers in South Korea have developed a multimodal deep learning model that fuses visual analysis and textual rubrics to automatically score drawing composition quality with expert-level accuracy.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results