Issues
- 0
box_nms: Boxes are negative values
#37 opened by shreyasr-upenn - 4
- 1
关于使用此superpoint代码训练后的测试效果?
#36 opened by PMRS-lab - 2
- 1
train the magicpoint with my own data
#32 opened by 00ohnim - 1
inference time
#31 opened by kaiyi98 - 0
how to test
#33 opened by Ma0zii - 15
Descriptor loss train
#24 opened by ericzzj1989 - 3
det_thresh in yaml files.
#29 opened by cxmscb - 8
请问训练sp时候的单应变换矩阵为什么要求逆呢?
#27 opened by dbdxzx - 1
Inference demo to get keypoints and descriptors?
#30 opened by cxmscb - 3
multi GPU train
#23 opened by RaneWouters - 2
About pre-trained model
#26 opened by kaiyi98 - 1
无法实现单机多卡训练
#28 opened by kaiyi98 - 1
about the label of 'homo_export_labels.py'
#25 opened by Onlymman - 3
About magic_point_coco_train.yaml
#22 opened by foxkw - 3
- 29
Training speed and results
#6 opened by litingsjj - 2
关于magicpoint是否只是用来生成coco label数据集?
#21 opened by atomishcv - 0
coco images size when superpoint trainning
#20 opened by sunwen-nreal - 3
Matching
#15 opened by Lucifer1002 - 3
README.md 中的 Usage 中是否有几个错误
#19 opened by ZhihangCui - 1
- 1
transfer learning
#14 opened by Lucifer1002 - 3
color input
#13 opened by zebin-dm - 1
train superpoint时的超参数设置
#11 opened by fish258 - 5
训练完成后使用时爆显存的问题
#10 opened by FeiXie8 - 1
Cuda: out of memory
#9 opened by hjr553199215 - 1
Superglue matching
#8 opened by Lucifer1002 - 1
- 1
Crop in augmentation_legacy.py
#5 opened by hjr553199215 - 3
- 2
Version requirements
#1 opened by SpiritAshes - 0
some envs problem
#3 opened by samadhiiiii - 2
RuntimeError: Expected 4-dimensional input for 4-dimensional weight [64, 1, 3, 3], but got 3-dimensional input of size [16, 120, 160] instead
#2 opened by yangninghua