-
I have tried it in cuda 11.0 torch1.9.0 but there are so many problems for it.
-
编译能够成功，但调用的时候总是出现类型不匹配的问题。
用的是windows平台。
ValueError: Buffer dtype mismatch, expected 'int_t' but got 'long long'
![image](https://user-images.githubusercontent.com/54108453/129647885-3fdbc134-32ce-…
-
Hai! Could someone please tell me how to solve this error. I am trying to train ReDet on HRSC2016 dataset. This is the error I am getting.
ReResNet Orientation: 8 Fix Params: False
2021-08-04 14:1…
-
## ❔Question
## Additional context
-
I looked into the DOTA website:
```
https://captain-whu.github.io/DOTA/dataset.html
```
I see Training , evaluation and testing data. How should I paste these data into the `datasets` folder …
-
您在论文里说"we prepare multi-scale data at three scales {0.5, 1.0, 1.5}, and random rotation for training and testing."
对于多尺度训练，您好似构建了一个多尺度的数据集而不是使用mmdet常规的方式，我猜测您的方式如下：将数据分别裁成512*512，1024*1024，1536*1536构…
-
I want to use the 79.42mAP model, but it only works in the old version. So I download the old one which need the .json annotation instead of .pkl . (If use the new version without Cascade, should I do…
Lg955 updated
3 years ago
-
I have a big problem and already tried many ways to find a solution for this error, but it still remains in my codes.
I am trying to reimplement the " BBAVectors-Oriented-Object-Detection" and firstl…
-
I 'm trying to use coco 's segmentation evaluation api for DOTA competition 's task 1. Are there any differences between DOTA_devkit 's IOU calculation (in `voc_eval` function) and coco 's IOU calcula…
-
##大佬再打扰一下，
1.repo里面在detect.py的时候会有一个NMS，这个NMS的过程如下图所示，所以这个NMS做完会出现一个物体被多个不同类别的框框住的现象；然后evalution.py中merge的时候也会有一个NMS，但是这个NMS对于每个物体只会保留一个置信度最大的框（不管是什么类别），这个理解对么？然后计算AP的时候就没有置信度阈值了？
![image](https://u…