Xiaoyu Zhang

I just tried using shared_ptr to solve this problem, [here](https://github.com/fishmarch/ORB-SLAM3-Remove-Bad).

> Probably because the rotate-center used in bevformer is not at (0, 0). > Any way, when adding tempral fusion(length-que = 3), I have a runtime error, have you got...

I was also puzzled with these codes. I think these are required because of mistakes in other codes. Such as, they use ego2global pose to retrieve map vectors, then they...

There may be some random for the final results. According to our experience, it is normal to get results around 66.0. How many ‘grad_norm: nan’ in your log? Too many...

Hi, sorry for the late reply. This may be the reason https://github.com/HXMap/MapQR/issues/19.

> Hi Xiaoyu, thanks for your reply! Yes I think this is the issue. Moreover, I also trained a MapQR model for 24 epochs (with the @auto_fp16() **[commended out](https://github.com/HXMap/MapQR/blob/main/projects/mmdet3d_plugin/bevformer/modules/encoder.py#L149)**), and...

Hi! Unfortunately, I didn't keep those logs. I can only provide a log training 'hrmapnet_mapqr_nusc_r50_24ep.py' recently. [hrmapnet_mapqr.log](https://github.com/user-attachments/files/17536912/hrmapnet_mapqr.log) Besides, I also notice the randomness of the reproduced results. I commonly get...

> > Hi! Unfortunately, I didn't keep those logs. I can only provide a log training 'hrmapnet_mapqr_nusc_r50_24ep.py' recently. [hrmapnet_mapqr.log](https://github.com/user-attachments/files/17536912/hrmapnet_mapqr.log) > > Besides, I also notice the randomness of the reproduced...

I haven't find any problem within the released codes. For current training strategy, the global map is generated for each epoch, then the model is trained with empty maps for...

Hi! Yes, it will certainly affect the global map and the final results. But the model has a certain robustness to such pose error when the error is not large,...