I was using the YoloSpatialDetectionNetwork node with a custom yolov8 model I made, but am getting this error. I am using Ubantu and OAK-d LR camera.

[INFO] [robot_state_publisher-1]: process started with pid [119042]

[INFO] [refuse_detection_node-2]: process started with pid [119044]

[robot_state_publisher-1] [INFO] [1723056437.989550522] [oak_state_publisher]: got segment oak-d-base-frame

[robot_state_publisher-1] [INFO] [1723056437.989595938] [oak_state_publisher]: got segment oak-d_frame

[robot_state_publisher-1] [INFO] [1723056437.989600452] [oak_state_publisher]: got segment oak_imu_frame

[robot_state_publisher-1] [INFO] [1723056437.989603360] [oak_state_publisher]: got segment oak_left_camera_frame

[robot_state_publisher-1] [INFO] [1723056437.989606165] [oak_state_publisher]: got segment oak_left_camera_optical_frame

[robot_state_publisher-1] [INFO] [1723056437.989608771] [oak_state_publisher]: got segment oak_model_origin

[robot_state_publisher-1] [INFO] [1723056437.989611423] [oak_state_publisher]: got segment oak_rgb_camera_frame

[robot_state_publisher-1] [INFO] [1723056437.989613970] [oak_state_publisher]: got segment oak_rgb_camera_optical_frame

[robot_state_publisher-1] [INFO] [1723056437.989616388] [oak_state_publisher]: got segment oak_right_camera_frame

[robot_state_publisher-1] [INFO] [1723056437.989618721] [oak_state_publisher]: got segment oak_right_camera_optical_frame

[refuse_detection_node-2] [2024-08-07 13:47:18.878] [depthai] [warning] [18443010E10D47F500] [1.12] Flashed bootloader version 0.0.26, less than 0.0.28 is susceptible to bootup/restart failure. Upgrading is advised, flashing main/factory (not user) bootloader. Available: 0.0.28

[refuse_detection_node-2] [18443010E10D47F500] [1.12] [2.386] [SpatialDetectionNetwork(4)] [warning] Network compiled for 6 shaves, maximum available 10, compiling for 5 shaves likely will yield in better performance

[refuse_detection_node-2] [18443010E10D47F500] [1.12] [5.610] [system] [critical] Fatal error. Please report to developers. Log: 'Fatal error on MSS CPU: trap: 09, address: 8009A710' '0'

Also what does "side6", "side13", and "side26" mean? What do the three digits associated with them mean? What is an anchor mask?

I also made the blob file via the BlobConverter CLI.

  • erik replied to this.

    bhavya_p Hi, did you get any crash report?
    https://docs.luxonis.com/software/depthai/examples/crash_report/

    The terms "side6", "side13", and "side26" are related to the anchor masks used in object detection models. In object detection models like YOLO, anchor masks are used to determine which anchors are responsible for prediction at each scale. The three digits associated with them are the indices of the anchors in the width, height, and depth dimensions respectively. Please consult google/GPT4 for such generic questions.

    Thanks, Erik

      it will not let me upload the json file the crash report creates. what type should i convert to

      @bhavya_p perhaps use gist, or upload to gdrive and send the link