User Guide 3: Simplifying Object Detection on a Hailo Device Using DeGirum PySDK

[Loi_Tran]

i added "Checksum": "", and my errror:

(degirum_env) RPiAI@Loitran:~/workspace/hailo_examples/py_hailo8l $ python test_detection.py 
Traceback (most recent call last):
  File "/home/RPiAI/workspace/hailo_examples/py_hailo8l/test_detection.py", line 4, in <module>
    model = dg.load_model(
            ^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/__init__.py", line 220, in load_model
    return zoo.load_model(model_name, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/zoo_manager.py", line 266, in load_model
    model = self._zoo.load_model(model_name)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/_zoo_accessor.py", line 623, in load_model
    assets = self._get_model_assets(model, True)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/_zoo_accessor.py", line 576, in _get_model_assets
    model_info = self._cloud_server_request(
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/_zoo_accessor.py", line 494, in _cloud_server_request
    raise DegirumException(details) from None
degirum.exceptions.DegirumException: 404 Client Error: Not Found for url: https://hub.degirum.com/zoo/v1/public/models/home/RPiAI/workspace/hailo_examples/pySDK/my_models/yolov11m/info
(degirum_env) RPiAI@Loitran:~/workspace/hailo_examples/py_hailo8l $ 

my JSON file:

{
    "ConfigVersion": 10,
    "Checksum": "",
    "DEVICE": [
        {
            "DeviceType": "HAILO8L",
            "RuntimeAgent": "HAILORT",
            "SupportedDeviceTypes": "HAILORT/HAILO8L"
        }
    ],
    "PRE_PROCESS": [
        {
            "InputType": "Image",
            "InputN": 1,
            "InputH": 640,
            "InputW": 640,
            "InputC": 3,
            "InputPadMethod": "letterbox",
            "InputResizeMethod": "bilinear",                
            "InputQuantEn": true
        }
    ],
    "MODEL_PARAMETERS": [
        {
            "ModelPath": "yolov11m.hef"
        }
    ],
    "POST_PROCESS": [
        {
            "OutputPostprocessType": "Detection",
            "PythonFile": "HailoDetectionYolo.py",
            "OutputNumClasses": 80,
            "LabelsPath": "labels_coco.json",
            "OutputConfThreshold": 0.3           
        }
    ]
}

Pls help me! i don’t know is "Checksum": "" or "Checksum": "SOMETHING"

[shashi]

Hi @Loi_Tran
Can you please confirm the contents of this folder? /home/RPiAI/workspace/hailo_examples/pySDK/my_models

[shashi]

Hi @Loi_Tran
Please enter any dummy value for checksum. The actual value does not matter as long as it is not left empty.

[Loi_Tran]

Here. Thanks very much!

[shashi]

Thanks @Loi_Tran for sharing this. Please add some dummy value to checksum and try again (do not leave the string empty).

[Loi_Tran]

Thanks @shashi for reply!

i was added dummy value to checksum but i get the same error:

(degirum_env) RPiAI@Loitran:~/workspace/hailo_examples/py_hailo8l/my_models $ ls
HailoDetectionYolo.py  labels_coco.json  yolov11m.hef  yolov11m.json
(degirum_env) RPiAI@Loitran:~/workspace/hailo_examples/py_hailo8l/my_models $ cd ..
(degirum_env) RPiAI@Loitran:~/workspace/hailo_examples/py_hailo8l $ python test_detection.py 
Traceback (most recent call last):
  File "/home/RPiAI/workspace/hailo_examples/py_hailo8l/test_detection.py", line 4, in <module>
    model = dg.load_model(
            ^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/__init__.py", line 220, in load_model
    return zoo.load_model(model_name, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/zoo_manager.py", line 266, in load_model
    model = self._zoo.load_model(model_name)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/_zoo_accessor.py", line 623, in load_model
    assets = self._get_model_assets(model, True)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/_zoo_accessor.py", line 576, in _get_model_assets
    model_info = self._cloud_server_request(
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/RPiAI/workspace/hailo_examples/degirum_env/lib/python3.11/site-packages/degirum/_zoo_accessor.py", line 494, in _cloud_server_request
    raise DegirumException(details) from None
degirum.exceptions.DegirumException: 404 Client Error: Not Found for url: https://hub.degirum.com/zoo/v1/public/models/home/RPiAI/workspace/hailo_examples/pySDK/my_models/yolov11m/info
(degirum_env) RPiAI@Loitran:~/workspace/hailo_examples/py_hailo8l $ 

I realy bad english! Thank for your help!

[shashi]

Hi @Loi_Tran
In your code you have model_zoo = '/home/RPiAI/workspace/hailo_examples/pySDK/my_models' but the model files seem to be located at '/home/RPiAI/workspace/hailo_examples/py_hailo8l/my_models'. Hope you can see the difference: pysdk instead of py_hailo8l.

[Loi_Tran]

Hi @shashi
I just wanted to say a big thank you for taking the time to help me fix that code bug. I was really stuck, and your support made a huge difference. I truly appreciate your patience and your willingness to jump in and assist.

[Loi_Tran]

By the way, I was thinking of trying this code using a camera instead of an image. Since I’m still new to Python, is there anything I should keep in mind or be careful about? I’d really appreciate any tips!

[shashi]

Hi @Loi_Tran
See our quick start guide on how to use it on video sources (video files, web cameras, rtsp stream): hailo_examples/examples/001_quick_start.ipynb at main · DeGirum/hailo_examples

[JanDev]

Hi, I have a finetuned Yolov8m model that I compiled to a hef file and i tried running it with Degirum SDK but got an error:

degirum.exceptions.DegirumException: [ERROR]Loading model failed
HailoRT Runtime Agent: Failed to create infer model, status = HAILO_FILE_OPERATION_FAILURE.
hailo_model_context_manager.cpp: 248 [DG::HailoModelContextManager::acquireModel]
When running model 'yolov8m'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/home/raspberry/Documents/hailo-rpi5-examples/basic_pipelines/detect.py", line 1270, in <module>
    main(args)
  File "/home/raspberry/Documents/hailo-rpi5-examples/basic_pipelines/detect.py", line 876, in main
    det_inf = detection_model(np.expand_dims(frame, 0))
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/raspberry/Documents/hailo-rpi5-examples/venv_hailo_rpi5_examples/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/raspberry/Documents/hailo-rpi5-examples/venv_hailo_rpi5_examples/lib/python3.11/site-packages/degirum/model.py", line 233, in call
    return self.predict(data)
           ^^^^^^^^^^^^^^^^^^
  File "/home/raspberry/Documents/hailo-rpi5-examples/venv_hailo_rpi5_examples/lib/python3.11/site-packages/degirum/log.py", line 59, in wrap
    return f(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^
  File "/home/raspberry/Documents/hailo-rpi5-examples/venv_hailo_rpi5_examples/lib/python3.11/site-packages/degirum/model.py", line 224, in predict
    res = list(self._predict_impl(source))
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/raspberry/Documents/hailo-rpi5-examples/venv_hailo_rpi5_examples/lib/python3.11/site-packages/degirum/model.py", line 1206, in _predict_impl
    raise DegirumException(
degirum.exceptions.DegirumException: Failed to perform model 'yolov8m' inference: [ERROR]Loading model failed
HailoRT Runtime Agent: Failed to create infer model, status = HAILO_FILE_OPERATION_FAILURE.
hailo_model_context_manager.cpp: 248 [DG::HailoModelContextManager::acquireModel]
When running model 'yolov8m'

shortened verion of the code:

detection_model = dg.load_model(
		model_name=args.degirum_model,
		zoo_url='models',
		inference_host_address='@local',
	)

cap = cv2.VideoCapture(0)
ret, frame = cap.read()
det_inf = detection_model(np.expand_dims(frame, 0))

Any ideas of what could the issue be?

[shashi]

HI @JanDev
Is this happening for only this model?

[JanDev]

Hi, I’m sorry for the message, I transferred the model using the local network and it got corrupted. Now it’s ok!

[JanDev]

Hi, sorry to interrupt again.
I was easily able to get the yolo model working, but Im having problems with an implementation of a zero_dce model. I want to enhance the captured image using this model and then feed the enhanced image to the yolo network.

I downloaded the enhanced model from this link (the compiled hef version; Btw I tested out the onnx version without degirum and it works) and wrote a script to test it out:

import degirum as dg
import cv2
import numpy as np

enhancement_model = dg.load_model(
    model_name='zero_dce',
    inference_host_address='@local',
    zoo_url='models',
)
cap = cv2.VideoCapture(0)
if not cap.isOpened():
    print("Error: Unable to open camera")
    exit(1)
cap.set(cv2.CAP_PROP_FRAME_WIDTH, 600)
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 400)

ret, frame = cap.read()
if not ret or frame is None:
    print("Error: Failed to capture frame")
    cap.release()
    exit(1)

cv2.imwrite("captured_image.jpg", frame)
print("Image saved as 'captured_image.jpg'")

frame_resized = cv2.resize(frame, (600, 400), cv2.INTER_LINEAR)
model_input = frame.astype(np.uint8)
enh_inf = enhancement_model(model_input)
enhanced_frame = enh_inf.image_overlay if hasattr(enh_inf, "image_overlay") else enh_inf.image
cv2.imwrite("enhanced_image.jpg", enhanced_frame)
print("Enhanced image saved as 'enhanced_image.jpg'")
cap.release()

But the saved enhanced image is exactly the same one that was captured before the enhancement (no change). What could be the problem?

[shashi]

Hi @JanDev
This is because the image_overlaymethod for these type of models is not supported. We are preparing an example on how to visualize the output image which we will share shortly.

[An_ti11]

Thank you very much for your reply. How can I set the batch size to 2? For example, I would like to just add two identical images. I would be very grateful for your answer.“inference_result = model(frame)
results = inference_result.results
box1 = np.array([det[‘bbox’] for det in results], dtype=np.float32)
score1 = np.array([det[‘score’] for det in results], dtype=np.float32)” “model = dg.load_model(
model_name=‘yolov11n_5’,
inference_host_address=‘@local’,
zoo_url=‘/home/zoo_url/yolov11n_5’
)”

[shashi]

Hi @An_ti11
Can you please explain what you mean by batch size of 2 and what adding two identical images means?

[Vlad_Klimov]

Hi, @An_ti11 ,
See my response here:

batch more than 1

Hi, @An_ti11 , You can use model.predict_batch() instead of mode.predict() to effectively pipeline a sequence of frames, see detailed description here: Running AI Model Inference | DeGirum Docs In a few words, you provide frame iterator as a method parameter, which, in turn, also returns iterator over results, which you can use in for loop: for result in model.predict_batch(["image1.jpg", "image2.jpg"]): Your input iterator may yield various frame types: strings containing image filenames n…