3detr

3detr copied to clipboard

Question about size loss.

1

In function loss_size, the "gt_box_sizes" is scaled by point_cloud_dims on the scale of 0 to ∞ while the "pred_box_sizes" is scaled into (0, 1) by sigmoid. F.l1_loss is used to...

Sharpiless

This may be a problem with incomplete uploaded code when running scripts/sunrgbd_ep1080.sh. This same as line4 in scripts/sunrgbd_ep1080.sh without a "\\".

Sharpiless

I wonder if you get 'nan' when using position embedding. Thanks.

liuhui0401

bbox_3d_iou, the same two 3d boxs, iou score is zero, zero.why?

4

utils > eval_det > box3d_iou. please   the box corners are generated by my_compute_box_3d function

JustDoItY

In their [recent paper](https://arxiv.org/abs/2203.10314) on a "Voxel Set Transformer", He et al. mention that 3DETR can only be applied to indoor datasets: > 3DETR present a promising solution by computing...

ghost

Update model_3detr.py

1

Minor updates to fix below error: File "~/repo/3detr/models/model_3detr.py", line 208, in run_encoder enc_inds = torch.gather(pre_enc_inds.type(torch.int64), 1, enc_inds) RuntimeError: gather_out_cuda(): Expected dtype int64 for index

chaoyivision

why do you perform down sampling after the first layer in 3detr-m, rather than the whole encoder?

2

Would this operation leads to performance drop? or because of the computational cost?

ch3cook-fdu

Problems with training on my own datasets

10

Dear author： At present, I have a batch of data sets to do 3D point cloud detection called center,which does not need to be classified, but only need to identify...

liudaxia96

Visualizing the attention

1

Hi, Do you plan to share the code for visualizing the encoder/decoder attention? Regards, Rishabh

theDragonSlayer97

Hi, thanks for the good research. I have a question about offset. Predicted offset value is in [-0.5, 0.5]. But when sampled query_xyzs have minimum distance 10 beween them, predicted...

JingukKim