[centerpose_decoder_node.NitrosSubscriber]: 10.0s passed while waiting to push a message entity (eid=1614) to the receiver camera_model_input #43
Comments
|
I encounter the similar problem, do you solve it? |
After I changed the system to ubuntu20.04 and JetPack5 version, everything was normal. I don’t know why it can’t run normally on ubuntu22.04 and JetPack6. |
Thanks, although it is a really strange problem |
I recently rebuilt the development environment of Isaac_ros and tried the centerpose algorithm again. This time I succeeded (successful within 10 minutes). If you need it, I have the following suggestions:
|
|
Did you find any solution for the multiple objects situation? I was trying to run the example given after installing and got the same problem since there are 2 shoes example |
Unfortunately, I haven't found a solution to the problem of multiple objects so far, but I have learned about the CenterPose algorithm itself. The algorithm does not seem to support the pose estimation of multiple objects. Of course, this is just my own judgment. If you want to solve this problem fundamentally, I think you can analyze it from the source code of CenterPose. I have been busy with other projects recently, so I have not further analyzed CenterPose. Regarding pose estimation, I finally used FoundationPose in my project. This algorithm has the best performance in terms of accuracy and stability that I have tested so far. Of course, if you are willing to continue to analyze the problem of the CenterPose algorithm, I look forward to your analysis and test conclusions. |
|
Hi @Yeager-101 I faced the similar issue as you. You mention of rebuilding using source code instead of apt installing. Do you mean i need to git clone
Best |
What I mean is that the centerpose package can be downloaded using git clone... and then manually compiled and tested, and other required packages can be directly installed using apt, if I remember correctly. |
So you mean, even though Best, |
What I mean is that you can try it this way. First, use apt to install all the function packages you need and their dependent function packages (including your centerPose function package), but you need to create a workspace in one of your folders, then git clone the Centerpose function package, and then compile it. After the compilation is complete, please modify the various paths under the launch file. Finally, please note that you should source your workspace source devel/setup.bash, and then roslaunch your centerPose node. I think there should be no more strange problems. |
|
Hi @Yeager-101 I found the solution to this kind of error of missing
Solution to this is shown below https://github.com/NVIDIA-ISAAC-ROS/isaac_ros_common/blob/release-3.0/docker/Dockerfile.aarch64 Mainly depend on the Best |
Congratulations on finding a solution to your problem. Sometimes CUDA, TensorRT, and Cudnn environments can cause a lot of errors. Although I didn’t encounter the problem you described when testing the package, I’m glad you can discuss the problem you encountered. |
|
Hello, I have encountered a similar problem as you and would like to ask for your advice. If possible, could you please add a vx or qq! Thank you very much! |
You can directly describe your problem here, so that it is easier for everyone to analyze the problem you encountered and better help others solve similar problems. |
Okay, thank you very much for your prompt response! Firstly, let me introduce my situation. I am using Ubuntu 22.04 and Jetpack 6.1 on Orin |
|
There are two more questions
|
I did not evaluate the frame rate of CenterPose, but the intuitive feeling is that it is less than 10FPS. As for how to use FoundationPose for pose estimation, there seem to be many reference projects on GitHub. |
Topic mapping is done using 'ros2 run topic_tools relay <source_topic> <destination_topic>', all you need to do is 'ros2 run topic_tools relay /camera/color/image_raw /image' and 'ros2 run topic_tools relay /camera/color/camera_info /camera_info' as I mentioned in my previous answer. |
I encountered an error after changing the topic of image information， [component_container_mt-1] 2025-01-13 14:22:54.547 ERROR /home/rx/workspaces/isaac_ros-dev/src/isaac_ros_dnn_inference/isaac_ros_gxf_extensions/gxf_isaac_tensor_rt/gxf/extensions/tensor_rt/tensor_rt_inference.cpp@543: Failed to retrieve Tensor input May I ask what is the reason for this? Can you help me check my isacc_ros_centerpose _tensor_rt. launch. py file to see if it has been modified correctly SPDX-FileCopyrightText: NVIDIA CORPORATION & AFFILIATESCopyright (c) 2021-2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.Licensed under the Apache License, Version 2.0 (the "License");you may not use this file except in compliance with the License.You may obtain a copy of the License athttp://www.apache.org/licenses/LICENSE-2.0Unless required by applicable law or agreed to in writing, softwaredistributed under the License is distributed on an "AS IS" BASIS,WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.See the License for the specific language governing permissions andlimitations under the License.SPDX-License-Identifier: Apache-2.0import os from ament_index_python.packages import get_package_share_directory def generate_launch_description(): |
When I try to use centerpose to start the node ros2 launch isaac_ros_centerpose isaac_ros_centerpose_tensor_rt.launch.py model_file_path:=/home/nvidia/Chen/centerpose/bottle_DLA34.onnx engine_file_path:=/home/nvidia/Chen/centerpose/bottle.plan, the terminal appears the following problem:
I don't know why this problem occurs. There is no image output in my "/centerpose/image_visualized" topic. My "/image" and "/camera_info" topics are normal.
The text was updated successfully, but these errors were encountered: