Replies: 1 comment 1 reply
-
|
作者还有兴趣吗？我正在做动态检测圣遗物框位置，目前是为了让playcover的兼容性更好一点。 另外个人感觉这种识别的目标其实很有限，未必需要深度学习法案，Bag-of-words应该就行。 |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
很欢迎动态框位置检测，bow可以试试，要是效果更好也能考虑 |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
👋 Yas未来计划支持扫描背包、角色等数据
但是作者并不熟CV，特来探讨可行的深度学习方案，或者传统CV方案
目前想到的背包识别方案：
如何提高CRNN准确率，或者其他的识别方案
目前是MobileNetV3_Small+LSTM
Beta Was this translation helpful? Give feedback.
All reactions