5 citations · 5 across the 1 of their papers we have counts for
1 paper
Jia Chen, Qin Jin
Sequence-level learning objective has been widely used in captioning tasks to achieve the state-of-the-art performance for many models. In this objective, the model is trained by t…