知识卡片
视觉问答联合图像与问题
VQA将图像和文字问题作为共同输入,输出答案,可用图像编码、问题编码和文本解码连接。需要以问题—图像—答案三元组训练或适配,并检查回答是否真正依赖图像;仅靠常见答案或问题先验也可能得到较高数据集分数。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 视觉与语言
原书出处 · 1 处
- · 查看原书依据
In this chapter, we have already seen all the tools necessary to create a VQA system; you just need a text encoder (for the question), an image encoder (for the image you are asking the question about), and a text decoder (for the answer). These can be hooked together using the same kinds of tools we have seen earlier in this chapter. The critical piece
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第51章《视觉与语言》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c051.xhtml)
- 结论依据:摘录中关于视觉问答联合图像与问题的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:In this chapter, we have already seen all the tools necessary to create a VQA system; you just need a text encoder (for the question), an image encoder (for the image you are asking the question about), and a text decoder (for the answer). These can be hooked together using the same kinds of tools we have seen earlier in this chapter. The critical piece