How does the YOLO test result return to the annotated txt file format?

[Aq114]

Search before asking

• I have searched the HUB issues and discussions and found no similar questions.

Question

I have conducted preliminary training on your platform and program to define the data set, and obtained some pictures of prediction results. However, I want to return the format of the annotation file from these results. What should I do?

Additional

There are categories for drawing borders and predictions in the test result picture.

[pderrenger]

@Aq114 hello!

Thank you for reaching out with your question. It's great to hear that you've been working with the Ultralytics platform and achieving prediction results! 😊

To convert the YOLO test results back into the annotated .txt file format, you can follow these steps:

1. Run Inference and Get Results:
   First, ensure you have the inference results in JSON format. You can use the Ultralytics HUB Inference API to get these results. Here's a quick Python example to get you started:

   import requests
   
   # API URL, use actual MODEL_ID
   url = "https://api.ultralytics.com/v1/predict/MODEL_ID"
   
   # Headers, use actual API_KEY
   headers = {"x-api-key": "API_KEY"}
   
   # Inference arguments (optional)
   data = {"size": 640, "confidence": 0.25, "iou": 0.45}
   
   # Load image and send request
   with open("path/to/image.jpg", "rb") as image_file:
       files = {"image": image_file}
       response = requests.post(url, headers=headers, files=files, data=data)
   
   results = response.json()

2. Extract and Convert Results:
   Once you have the results, you can extract the necessary information and convert it into the YOLO annotation format. The YOLO format typically includes the class, center x, center y, width, and height of the bounding box, all normalized by the image dimensions.

   Here's a sample code snippet to convert the results:

   import json
   
   def convert_to_yolo_format(results, image_width, image_height):
       annotations = []
       for detection in results['data']:
           class_id = detection['class']
           x_center = detection['xcenter'] * image_width
           y_center = detection['ycenter'] * image_height
           width = detection['width'] * image_width
           height = detection['height'] * image_height
           annotations.append(f"{class_id} {x_center} {y_center} {width} {height}")
       return annotations
   
   # Assuming you have the image dimensions
   image_width = 640
   image_height = 480
   
   yolo_annotations = convert_to_yolo_format(results, image_width, image_height)
   
   # Save to a .txt file
   with open("path/to/annotations.txt", "w") as file:
       for annotation in yolo_annotations:
           file.write(f"{annotation}\n")

3. Verify and Adjust:
   Ensure that the image dimensions (image_width and image_height) are correctly set to match your input image. The annotations will be saved in the YOLO format in the specified .txt file.

If you encounter any issues or if the problem persists, please make sure you are using the latest versions of the Ultralytics packages and the YOLO model. Updates often include important bug fixes and improvements.

For more detailed information, you can refer to the Ultralytics HUB Inference API documentation.

Feel free to reach out if you have any further questions. Happy coding! 🚀