Assistant Professor, Johns Hopkins University
1 paper at NeurIPS 2025
A training-free, generative approach that infers object removal order by exploiting statistical co-occurrence and asymmetry priors learned by generative models.
