yatengLG/Retinanet-Pytorch
Retinanet目标检测算法(简单,明了,易用,全中文注释,单机多卡训练,视频检测)(based on pytorch,Simple, Clear, Mutil GPU)
Python
Issues
- 0
ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (6,) + inhomogeneous part.
#28 opened by xyj176 - 1
- 0
visdom server 網頁無法開啟
#27 opened by aaa157554 - 2
报错
#21 opened - 1
VOC数据集的最终mAP
#26 opened by wwwbobo - 2
fpn最后的3X3卷积都是用的conv1的吗?
#25 opened by Project627 - 1
方便把fpn 的 p6, p7 加回来吗?
#24 opened by bjmwang - 7
RunTimeError
#22 opened - 14
为什么会出现这个错误RuntimeError: Expected object of scalar type float but got scalar type double for argument 'other'
#17 opened by jfdyft - 2
false positives 特别多的问题
#20 opened by xzxedu - 0
欢迎大家star、fork
#19 opened by yatengLG - 2
- 0
检测速度问题
#7 opened by Complicateddd - 0
- 3
我训练的数据集,目标比较小,训练的时候loss很低,但是检测不出来,ap值三点几
#16 opened - 1
如何分别设置输入图片IMAGE_SIZE的长宽?
#13 opened by ZhanJunxiang - 1
- 10
输入尺寸和预测框大小
#9 opened by Hongzhehao - 4
网络如何进行微调和迁移学习
#12 opened by mur909 - 1
BUG! 当图片(.xml)中不包含任何 object 时!
#11 opened by i4yyds - 0
- 0
- 2
关于计算loss相关
#5 opened by pangkun248 - 2
assign_priors里面负样本相关
#6 opened by pangkun248 - 3
可不可以放几张效果图呢?或者在coco数据集上的map
#4 opened by yangbisheng2009 - 11
net.load_pretrained_weight 没这个方法呢?
#2 opened by yangbisheng2009 - 1
detect测试,很多框都是重合的,是不是没有nms?
#3 opened by yangbisheng2009 - 2