Toybrick

有人成功跑通了yolov2吗?

guigen80

新手上路

积分
31
发表于 2019-3-26 14:31:22    查看: 6299|回复: 4 | [复制链接]    打印 | 显示全部楼层
我用的caffe版本试的,缺少reorg层的支持。我试着用我们剪掉的版本,发现无法标出物体。有人跑通过吗?谢谢
回复

使用道具 举报

chuyee

中级会员

积分
352
发表于 2019-3-27 02:37:21 | 显示全部楼层
I tried the Yolov2 from darknet. It works great on rknn. Any specific reason do you want to use the caffe version of Yolov2?
回复

使用道具 举报

guigen80

新手上路

积分
31
 楼主| 发表于 2019-3-27 09:29:19 | 显示全部楼层
yes, we train it in caffe, it can't support reorg layer. can we add the custom layer for rknn?
回复

使用道具 举报

chuyee

中级会员

积分
352
发表于 2019-3-27 13:08:07 | 显示全部楼层
guigen80 发表于 2019-3-27 09:29
yes, we train it in caffe, it can't support reorg layer. can we add the custom layer for rknn?

I'll leave RK people to comment how to add custom OP support for rknn. Personally I won't think it will be easy even if it is supported.

Why do you think rknn doesn't support reorg layer? I don't think I had any issue when converting yolov2 (which contains reorg layer) from darknet.
回复

使用道具 举报

guigen80

新手上路

积分
31
 楼主| 发表于 2019-3-27 14:21:09 | 显示全部楼层
T google.protobuf.text_format.ParseError: 915:5 : Message type "caffe.LayerParameter" has no field named "reorg_param".

when to transform , it report the error, so I think it can't support reorg layer. I use the model from yolo9000
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则

产品中心 购买渠道 开源社区 Wiki教程 资料下载 关于Toybrick


快速回复 返回顶部 返回列表