Haiwen Li

Peak-End-Net: A Peak-End Rule Inspired Framework for Generalizable Video Aesthetic Assessment

Geng Li, Haiwen Li, Rui Chen, Jing Tang, Lei Sun, Xiangxiang Chu

Video aesthetic assessment (VAA) aims to predict how aesthetically pleasing a video is, yet remains far less explored than other visual assessment tasks. Its progress is hindered not only by the scarcity of large-scale benchmarks, but also by the intrinsic subjectivity of aesthet…Read more