Issues
- 0
Couldn't find depth loss function
#20 opened by bochennn - 22
Low mAP and NDS performance when eval camera stream full-training checkpoint with lower input size.
#28 opened by lufanma - 4
- 7
数据预处理时遇到问题
#18 opened by lihuashengmax - 3
你们是怎么训练的
#15 opened by rubbish001 - 7
- 0
- 0
Can you provide pre-trained models mask_rcnn_dbswin-t_fpn_3x_nuim_cocopre.pth?
#29 opened by Da1symeeting1 - 6
loss is too small
#23 opened by hehern - 4
Memory Requirements - Reg
#22 opened by jyothsna-phd22 - 6
Code based on bevfusion-mit
#16 opened by think-twice-1218 - 0
- 1
ealss怎么加入到bevdepth里,求作者大大解答
#19 opened by doudoudou12 - 0
- 0
有A100赞助一下吗
#26 opened by rubbish001 - 0
- 0
Installation problem
#17 opened by DodgePark - 5
数据生成问题
#14 opened by jianpingZhonggit - 0
About visualisation of depth map.
#13 opened by YangChen1234567 - 1
TTA + ensemble procedure
#11 opened by zlenyk - 1
- 5
Test time augmentations
#8 opened by anuj-sharma-19 - 4
Segmentation results
#9 opened by ldtho - 4
- 0
About the dimension of depth
#7 opened by jjlinghu - 2
Nuscenes detection errors on validation set
#5 opened by donand - 10
BEVDepth Implementation
#2 opened by Young98CN - 1
- 7
what's the accuracy of camera pretrained model, which use depth map as input
#3 opened by BoomSky0416 - 3
Great work,expect code to be released.
#1 opened by JsutCheng