Self-critical Sequence Training for Image Captioning.
Computer Vision and Pattern Recognition（2016）
关键词
reinforcement learning,MSCOCO task,self-critical sequence training,SCST,test-time inference algorithm,reward signal,actor-critic methods,REINFORCE algorithms,MSCOCO evaluation,policy-gradient methods,deep end-to-end systems,image captioning
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要