清晨的雾

Hello @Four1996 I'm training on a custom dataset and it seems that it costs nearly a week to complete 80 epochs. I just modified the annotation file path in `yolox_x_mix_det.py`...

In my log the `data_time` can be up to 30s, I guess that may be the reason. I'm trying to solve it.

Having moved the data to SSD, the problem has been solved.

@ckqsars In fact, in CenterNet on which the detection branch is built upon, the authors said that max is used

> Looking at the bounding boxes it seems, you're not taking care of the offset, prediction can be center(x, y), width and height. > So to get top-left coordinate you...

I have noticed this update: https://github.com/google-research/kubric/tree/main/challenges/movi#movi-e

Hello, In inference, the `get_eval_loader` function defined in the exp file is used to get the DataLoader: https://github.com/MCG-NJU/MixSort/blob/a078f5bf6ae9fbeecbc1384479d5f02ab8b9e7f6/tools/track_mixsort.py#L175-L183 Since we haven't provided a script to directly run inference on custom...

Hi, From the error message you provided, it seems that the `template_imgs` variable is not a list of `torch.Tensor` objects as expected. The error indicates that one or more elements...

@SergeySandler From your details, the issue seems to originate from the exception handling in https://github.com/MCG-NJU/MixSort/blob/52d317c1c79860afe705dd2252c4ae754308b75e/yolox/mixsort_oc_tracker/mixformer.py#L66-L72 This behavior is actually a simplified adaptation from the handling below where data is marked...

Hi @OOHAN-zxy , These parameters are set using this function: https://github.com/MCG-NJU/MixSort/blob/a078f5bf6ae9fbeecbc1384479d5f02ab8b9e7f6/MixViT/lib/train/base_functions.py#L11-L24 and the `cfg` is loaded from `MixViT/experiments/mixformer_deit/track.yaml`