Maybe it's not too significant,

but it seems like your Bilateral Context Module ends with DownSampling, whereas the figure 2. in your paper describes BCM to end with Bilateral context block. Is it intended or is there something I'm missing?

Hi, it is a good job.I am interested in it! But I am confused with aug_loss. The code is followed:

aug_loss_weights = tf.constant([0.1, 0.1, 0.3, 0.5, 0.5]) aug_loss = 0 # new_xyz_list=（B,N,16,3）...xyz_list=（B,N,3)... for i in range(self.config.num_layers): centroids = tf.reduce_mean(self.new_xyz[i], axis=2) # (B,N,3) relative_dis = tf.sqrt(tf.reduce_sum(tf.square(centroids-self.xyz[i]), axis=-1) + 1e-12) # （B,N,1） aug_loss = aug_loss + aug_loss_weights[i] * tf.reduce_mean(tf.reduce_mean(relative_dis, axis=-1), axis=-1) #weight*B

Q1: why is the aug_loss_weight assigned like this？ Q2: the output of code" tf.reduce_mean(tf.reduce_mean(relative_dis, axis=-1), axis=-1)" is B? Is my remarks correct? I hope to get your reply.

Dear Sir: Hi , I am confused with the shifted problem in the bilateral context module.According to my memory, the RandLA-Net is not have the shifted problem.And your code is built on RandLA-Net.Is this a basic challenge in the point cloud semantic segmentation? what caused you to consider to solve it？I would like to take this opportunity to learn your way of thinking from you.Moreover, can you share your opinions about the problems exited in point cloud semantic segmentation.

Dear sir: Have you try to use multi-gpu for your training? I have changed the code os.environ['CUDA_VISIBLE_DEVICES'] = "str(FLAGS.gpu)" to os.environ['CUDA_VISIBLE_DEVICES'] = "0, 1" When I type the message "nvidia-smi" in the command line , and I found that gpu1 is still not working in fact. It still have the problem of "OOM".How should I fix this?

Hello,

I have been doing projects on implementing your code on PyTorch, and the link is here

https://github.com/deepshwang/BAAF-pytorch

However, the model I coustructed is not converging, and is in process of debugging it...

I know it's a tedious and time-consuming to review other's code, but wish you may review the model if you have some free time.. :)

Many thanks to your outstanding work!

It is a good job, but I encountered some problems. After processing the data, there is no error, and the .ply file is also generated, but when I run BAAF-Net.py and main_s3dis.py , both files report the same error: 'process finished with exit code 139 (interrupted by signal 11: SIGSEGA)'.Besides, It can't Debug in the Pycharm. This troubles me. What should I do in this situation?

Hi, @ShiQiu0419 I am very interesting in you work, so i want to know when the source code of this project will be published? Thank you very much!

Best Regards.

Hi, Dear PhD.QIU. @ShiQiu0419 Thanks for your opensource code. And I have some question need help.

• Since the open source code does not contain the SemanticKITTI dataset, I coded and trained this part based on the RandLA-Net code and it seems to work fine. However I am not sure if there is a potential problem, can you help me review this section? This part of my code is in https://github.com/huixiancheng/My_BAAF_with_SemanticKITTI

• I note that you conducted comparative experiments on sampling methods in this part of the ablation experiments in the supplementary material. The repo is the official code, so it does not include test part of the ablation experiment. Currently I am learning the differences between different sampling methods, so I want to implement BAAF with RS. Which means bigger modification, mainly including data batch process parts like below. https://github.com/ShiQiu0419/BAAF-Net/blob/663d1681d4d05ad3caaacd98e6dedfdc9caa4930/main_S3DIS.py#L168-L177 Since I am not good at tensorflow framework pipeline, can you please share this part of the code? My email address is [email protected].

Thank you in advance for any potential help.

EPOCH 0 Step 00000050 L_out= nan Acc=0.00 --- 1006.39 ms/batch Step 00000100 L_out= nan Acc=0.00 --- 924.97 ms/batch Step 00000150 L_out= nan Acc=0.00 --- 908.08 ms/batch Step 00000200 L_out= nan Acc=0.00 --- 760.66 ms/batch Step 00000250 L_out= nan Acc=0.00 --- 811.27 ms/batch Step 00000300 L_out= nan Acc=0.00 --- 967.48 ms/batch Step 00000350 L_out= nan Acc=0.00 --- 719.22 ms/batch Step 00000400 L_out= nan Acc=0.00 --- 828.67 ms/batch Step 00000450 L_out= nan Acc=0.00 --- 696.24 ms/batch Step 00000500 L_out= nan Acc=0.00 --- 873.87 ms/batch Step 00000550 L_out= nan Acc=0.00 --- 719.05 ms/batch Step 00000600 L_out= nan Acc=0.00 --- 788.21 ms/batch Step 00000650 L_out= nan Acc=0.00 --- 798.09 ms/batch Step 00000700 L_out= nan Acc=0.00 --- 925.55 ms/batch Step 00000750 L_out= nan Acc=0.00 --- 741.91 ms/batch 0 / 200 50 / 200 100 / 200 150 / 200 eval accuracy: 0.18944735107421876 mean IOU:0.01457287315955529 Mean IoU = 1.5%

1.46 | 18.94 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Best m_IoU is: 1.457

I want to Print model summary, but when I call model.summary() in s3dis.main after model = Network(dataset, cfg) attribute error occurs. If you know any other way to calculate the number of model parameters that will be also helpful.

thankyou

Error: AttributeError: 'Network' object has no attribute 'summary'

Hi,thanks for your @ShiQiu0419 Due to I saw the network have a parameter "num_points" in network to do Fathest Point Sampling,so why not use point.shape[0] as a dynamic n_points? Because the number of point in semantic KITTI is not all the same but in a range of 120000~130000.So whether the performance will be influenced if I change the way of setting the number of point?

I have a question about how the network be used on semantic KITTI that if the input is 64*2^10 with batch_size 1, the gpu memory useage is about 18G+ like below:

Sat Dec 11 21:38:27 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 465.31       Driver Version: 465.31       CUDA Version: 11.3     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  Off  | 00000000:01:00.0  On |                  N/A |
| 36%   47C    P2   114W / 370W |  18763MiB / 24265MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A      1005      G   /usr/lib/xorg/Xorg                596MiB |
|    0   N/A  N/A    723501      C   ...conda3/envs/pc/bin/python    18163MiB |
+-----------------------------------------------------------------------------+

Thanks for your excellent work about point cloud semantic segmentaton。 I have tested the S3DIS dataset with the code you shared。But my own dataset is same as semantic3D dataset，so I want to know if you have any plans to share the code for training the Semantic3D dataset recently。

Hello, I have followed your tips, and trained the model. On Area 5, S3DIS dataset, the model I trained only achieved 61.078 mIOU. The mIoU in your paper is 65.4. Could your share your training parameters on S3DIS dataset? Looking forward to your reply.