Hi,
I tried sane annotation tool in my Ubuntu 18.04 system. There is no GPU. After running python app.py on the terminal, opened the browser, selected the pcd file, but point cloud is not uploaded. On closing the terminal only the point cloud is uploaded in the browser. For annotating next frame, again python app.py is ran on the terminal, opened the browser, selected the frame, closed the terminal and point cloud is uploaded. Why it is so?

Hey @hasanari,

I am getting above errors on Windows. Is there any solution to solve this or load the point clouds without GPU?
I have GPU installed on my machine and it is recognised successfully.

It would be great if you could suggest some possible fix to overcome this issues.

Hi,

From the demo gif, I am not able to find any option to adjust the height of the bounding box. It looks like the tool is predicting the height by itself, but what if I would like to change the height?

I am running the tool with all the the requirements installed along with tensorflow-gpu=2.4.0, However, I get an error
for the following line

Placeholders

indices = tf.placeholder(tf.int32, shape=(batch_size, None, 2), name="indices")
is_training = tf.placeholder(tf.bool, name='is_training')
pts_fts = tf.placeholder(tf.float32, shape=(batch_size, max_point_num, setting.data_dim), name='points')
######################################################################

It seems the file "predicting_point_segmentation.py" still uses a functionality of tensoflow 1.X .
Could you please suggest a fix?