gnayuohz / centernet-visdrone Goto Github PK

哥们，我在mmdetection里面用centernet_resnet18跑visdrone在val上得到了29.4的map，比你用resnet50还高，挺奇怪的

您好，我是一个初学者，正如题目所说，我在结合您这个代码和官方版https://github.com/xingyizhou/CenterNet 一起在看代码学习，我用您这个跑出来的模型，想试试在官方版里试试推论，但是参数和代码不知道怎么改，请多多指教！

Hello. I am wondering that I am trying to train the 1 class object detection using your code.
And then there is error bout
for iter_id, batch in enumerate(train_loader):
Exception has occurred: IndexError
Caught IndexError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/henry/miniconda3/envs/flowtorch/lib/python3.8/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/home/henry/miniconda3/envs/flowtorch/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 49, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/henry/miniconda3/envs/flowtorch/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 49, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/henry/centernet-visdrone/src/dataset.py", line 185, in getitem
np.maximum(gs_mask, hm[i], out=gs_mask)
IndexError: index 1 is out of bounds for axis 0 with size 1
File "/home/henry/centernet-visdrone/src/dataset.py", line 205, in
for iter_id, batch in enumerate(train_loader):

any suggestion?

Hello, I want to ask you, when you train the model with Visdrone dataset, how long does it take to do the training?

Hello, I wanna ask you, did you resize all of the images to 1280*960 ??? because I look in to the dataset, they all have different size on the training set visdrone.

Thank you!