Image-based virtual try-on is one of the most promising applications of humancentricimagegeneration duetoitstremendous real-worldpotential.

We introduce the problem of Bayesian optimization with cost-varying variable subsets (BOCVS) where in each iteration, the learner chooses a subset of query variables and specifies their values while the rest are randomly sampled.

Optimization theory is a central research topic in computer science, mathematics, operations research, etc. Currently, many efficient algorithms for optimizing convex functions have been proposed (see for instance [10]), but much less is known for nonconvex optimization.

Annotateddimensionsinclude Valence, Arousal and Emotion label, with annotations gathered using Prolific.