Liaoqing-up

handle different numbers of point cloud

3

hello, i want use PU-Net to upsample point cloud from robosense lidar16, the number of points in each frame maybe different, but the source code show that the **tf.placeholde**r must...

Model: pretrained model

1

can you provide some pretrained model and config, like '**R_101_FPN_2x_nucoco_it30000_fb/model_final.pkl**' and '**R_101_FPN_2x_nucoco_it30000_fb/fast_rcnn_R-101-FPN_2x_nucoco.yaml**' mentioned in **3_test.sh**, thankyou!

why the 'max_num_pts_per_box'

1

https://github.com/Abyssaledge/TorchEx/blob/e5ff7fae27b5b6df74e6204dac45554ba85117a2/torchex/src/dynamic_point_pool/dynamic_point_pool_kernel.cu#LL297C41-L297C41 I'm confused about this variable and this code. Can you explain it? Thank you~

Hello, tianweiy. I would like to ask a question that has been bothering me for a long time. The common practice in the nusecens dataset is to divide the ten...

Question about why the add&norm structure of the tranformer network differ from the typical transformer one

3

https://github.com/TuSimple/centerformer/blob/96aa37503dc900d1aebeb7c1086c33bbd0c01d26/det3d/models/utils/transformer.py#L267-L279 In the code, the residual in transformer is only the input after **add** and does not pass through the **norm** layer. add and norm are not taken as a...

some questions about nuscenes multi-task support

2

Thanks for releasing the nuscenes dataset code support. I have some questions about the implement of the multi-tasks. I see in the code that you define obj_num=500 for each task...

Some details to discuss

1

Thank you for open-sourcing your work. I was wondering, why you use x_up(the current frame's bev feature) other than x_up_fuse(the sequential frames through spatial-aware fusion) as center query embedding ?...

Thank you for open-sourcing your work. I was wondering about the image-guided query initialization, it seems the image collapse feature only used to generate heatmap, but not used for the...