CRF-RNN
2000년 도입 · 논문 4편에서 사용
CRF-RNN is a formulation of a CRF as a Recurrent Neural Network. Specifically it formulates mean-field approximate inference for the Conditional Random Fields with Gaussian pairwise potentials as Recurrent Neural Networks.
출처: Conditional Random Fields as Recurrent Neural Networks
소개 논문: Conditional Random Fields as Recurrent Neural Networks
Recurrent Neural Networks · Sequential