Issues
- 0
关于分割指标IoU和mIoU的计算?
#78 opened by Cxgmg - 2
可行驶区域
#51 opened by worlkingbling - 4
如何增加一个分割任务
#60 opened by PotatoNU - 6
train.cache
#52 opened by worlkingbling - 0
请问自己训练模型时,如何修改输入图片的尺寸?
#77 opened by wycrystal - 0
执行python predict.py 遇到问题
#76 opened by wycrystal - 25
- 7
- 1
Cannot download image data
#69 opened by zhang-pan - 0
- 2
- 2
pixacc refers to
#74 opened by jasfa - 2
- 0
ValueError: wandb.log must be passed a dictionary
#73 opened by Reversev - 2
Tensor RT conversion
#63 opened by Venkat-1405 - 4
- 1
請問如何印出每一層的FLOPs?
#68 opened by s2244521 - 1
map: [None,{'10':'0'},{'11':'0'}]
#62 opened by oyoanan - 1
RuntimeError: The size of tensor a (8000) must match the size of tensor b (8400) at non-singleton dimension 1
#64 opened by 1402091247 - 1
关于收敛问题?
#66 opened by worlkingbling - 0
- 0
谢谢作者分享!我想请问如何在v8中加入miou的评价指标?
#65 opened by handsome2456 - 12
你好,我下载images的路径打不开:https://bdd-data.berkeley.edu/
#39 opened by mco0308 - 2
为什么识别结果的标识都是"Person"
#61 opened by oyoanan - 4
Do this code has ONNX and Tensor RT support
#28 opened by Venkat-1405 - 1
想问一下各位大佬,车道线的结果如何才能提取并使用呢?
#56 opened by QiTianDaShengDaShi - 1
我使用result=model.predict(source=input, device=[0],name='0516', save=True, conf=0.25, iou=0.45, show_labels=True)代码推理,但是返回值result没有结果,但是官方yolov8使用相同的result=model.predict 就有返回值,该处控制是否有返回值的代码在具体哪里啊
#50 opened by QiTianDaShengDaShi - 1
目标检测类别有挺多,但是实际跑的时候只能检测车辆
#55 opened by QiTianDaShengDaShi - 0
从images下载图像打不开
#58 opened by wangwangzuran - 2
没有看到预训练模型呢
#57 opened by oyoanan - 7
關於Precision很低的問題
#43 opened by s0966066980 - 1
求助 这个是怎么回事 所以数据集的目录都是对的 还显示这个错误
#54 opened by worlkingbling - 2
请问一下 这个错误是怎么回事啊?
#53 opened by worlkingbling - 2
bdd-multi.yaml
#47 opened by taotaoland - 1
这个问题怎么回事
#49 opened by worlkingbling - 5
你好我想问下如果训练速度太慢是因为我的图片大小不是(720,1280)的问题么。
#48 opened by qingtian787 - 3
- 1
训练很慢是正常的吗?比常规的YOLOv8训练慢太多了,cache=True还是很慢
#45 opened by Tomhardy13 - 1
关于mIoU
#41 opened by YangBo0411 - 1
关于视频预测后保存问题
#44 opened by libaos - 6
det+line_seg
#30 opened by mengzhaoZ - 1
Why does the Adaptive Concatenation Module choose 0.5 as the threshold for sigmoid(weight)?
#37 opened by TonyMacedonia - 1
- 4
- 0
own dataset
#42 opened by YangBo0411 - 6
own datasets
#36 opened by s0966066980 - 2
model pruning
#35 opened by loganwu0526 - 1
- 1
我将conv里面的激活函数变成了relu,然后训练模型,显示IOU=0
#31 opened by pengyang1225 - 0