Seung-Kyum Choi

The paper reframes industrial visual sim-to-real transfer as a domain-gap problem categorized by the availability of explicit object geometry (CAD), arguing that the required prior evidence dictates the appropriate deployment strategy.