- Edited
To use YoloDetectionNetwork, you can try this
To use YoloDetectionNetwork, you can try this
We can see that there is a YoloDetectionNetwork node to decode the output of yolo.
I have some yolov3/yolov4 models, they are built by different deep learning frameworks(such as pytorch, tensorflow...), and their output is a bit different. I want to use them on Oak, and I want to use the YoloDetectionNetwork node to parse the output.
What I want is what model output the YoloDetectionNetwork node needs, so that I can adjust the existing model to fit the YoloDetectionNetwork node.
depthai/depthai_helpers/calibration_utils.py
I think you can modify this file by adding any code at line 323.
For example,
np.savez('calib.npz', Q=self.Q)
Then perform a calibration according to the document, and the files you need should be saved to the depthai/depthai_helpers folder
Okay, thank you for your reply.
II want to use the second generation pipeline builder to display the depth information of the detected object, but the depth map is not aligned with the color map, there is a deviation.
Is there a script for aligning depth maps with color maps ?
This my code:
Object_Detection
Recent versions of DepthAI Pipeline Builder Gen2 have video delays.
On the Ubuntu 18.04 system, I tested using mobilenet-ssd for object detection.
These are some of the versions I've tried.
0.0.2.1+b15a4a1740fef3d805634eaefbcb8ce136fc2d26.
0.0.2.1+b9918c481abaaeac51dd1d2b80fe6a150612f90a.
0.0.2.1+bc6a6c484c7e805659b83c091fe91c7c4ace22c9.
0.0.2.1+57205e4dad905895090ef51237deea0905390e66.
Using the above version, I move the OAK-D and the display will be delayed by a few seconds.
The delay in this old version is not so obvious.
0.0.2.1+83f365852d249ac72fc68ad6c40cc9745994c115.
Using this version, I move the OAK-D and display it in almost real time.