请问数据集是如何处理的呢

[ARONZXA]

校准文件缺少，只有一个单独的calib.txt

[liujiuming123]

Thanks! We have modified the file config.py now and you can check this time.

[Smezzdl1996]

校准文件缺少，只有一个单独的calib.txt

I have the same issue! when i run the train.py there are something mistake that the system can t fine the calib.txt per frame as:
FileNotFoundError: [Errno 2] No such file or directory: './data_odometry_color/dataset/sequences/07/calib/000000.txt'
is there any additional process to convert the single calib.txt to each frame?

[liujiuming123]

It is directly downloaded from raw data in KITTI website, but with a conprehensive one.

[bkk12]

How to provide a separate calibration file for each frame or sensor in the dataset, rather than a comprehensive file

[TYBHC0052]

校准文件缺少，只有一个单独的calib.txt

我有同样的问题！当我运行 train.py 时，系统无法对每帧的calib.txt进行微调，因为：FileNotFoundError：[Errno 2] 没有这样的文件或目录：'./data_odometry_color/dataset/sequences/07/calib/000000.txt' 是否有任何额外的过程可以将单个calib.txt转换为每一帧？

May I ask if this issue has been resolved? I want to ask for advice.

[hhxin2001]

It is directly downloaded from raw data in KITTI website, but with a conprehensive one.
In addition to the problem with the dataset, I run this code now, and there is also a problem with dimension inconsistencies in the network part, what is the reason for this?
Traceback (most recent call last):
File "train.py", line 341, in
main()
File "train.py", line 134, in main
eval_pose(model, test_dir_list, init_epoch)
File "train.py", line 276, in eval_pose
model(pos2, pos1, imback2, imback1, xy2, xy1, T_gt, T_trans, T_trans_inv, pts_valid2, pts_valid1)
File "/home/sdb/hehaixin/.conda/envs/dvlo/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/home/sdb/hehaixin/.conda/envs/dvlo/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
return forward_call(*args, **kwargs)
File "/home/sdb/hehaixin/DVLO/pwclonet_model.py", line 332, in forward
l0_xy_cor_f1[batch, :l0_batch_xy_f1.shape[0], :] = l0_batch_xy_f1 # [B, N, 2] 当前批次有效点的二维坐标
RuntimeError: The expanded size of the tensor (700) must match the existing size (1381) at non-singleton dimension 0. Target sizes: [700, 2]. Tensor sizes: [1381, 2]

[diuzhevVlad]

@liujiuming123 Hi, thank you! Do you have any updates on this issue?

[Lyoid]

求求更新一下这个祖传的readme吧

[Tian0822]

校准文件缺少，只有一个单独的calib.txt

我有同样的问题！当我运行 train.py 时，系统无法对每帧的calib.txt进行微调，因为：FileNotFoundError：[Errno 2] 没有这样的文件或目录：'./data_odometry_color/dataset/sequences/07/calib/000000.txt' 是否有任何额外的过程可以将单个calib.txt转换为每一帧？

May I ask if this issue has been resolved? I want to ask for advice.

@TYBHC0052 Hello, has this problem been solved? I also encountered this problem.