RFCN

RFCn is an acronym that can refer to more than one concept, depending on the field. In the context of Internet standards, RFCn most often denotes the nth document in the Request for Comments series published by the Internet Engineering Task Force (IETF). RFCs are publication identifiers for documents that describe internet protocols, procedures, and related information. They are numbered sequentially; the series began in 1969, and RFCs cover topics from foundational protocols to informational memos. Not every RFC is a standard; many are informational, experimental, or best-current-practice documents. Each RFC indicates its status and applicability, and RFCs are maintained in the RFC Editor archive and associated registries.

In computer vision and machine learning, RFCN is used to refer to Region-based Fully Convolutional Networks, an approach to object detection. The idea is to perform region-based detection using a fully convolutional backbone to share features across candidate regions, aiming to improve efficiency relative to earlier R-CNN approaches. In practice, the term is sometimes written as RFCN, while the more common abbreviation for this approach is R-FCN; both refer to related families of methods for object detection.

Because RFCn is ambiguous, context is essential to determine which interpretation is intended. See also IETF, RFC, R-FCN, object detection.