Browse categories
Explore
Fiverr Pro
English
$
USD
I will do polygon annotation, image label and bounding box for computer vision datasets
Python Developer: Web Scraping, Automation and YOLO Computer Vision Expert
Python developer building computer vision systems, automation tools, and sports bet platforms. 7+ years writing production code.
Daily work: YOLO and OpenCV detection, tracking and counting apps. Data...
About this Gig
Are you looking for pixel-perfect data annotation and image labeling to train your computer vision models? Look no further!
Welcome to a professional, high-accuracy data annotation and image labeling service. As a specialized computer vision assistant, I understand that the accuracy of your AI model entirely depends on the quality of your dataset. I provide meticulously labeled data that is ready for instant training in YOLOv8, YOLOv10, PyTorch, or TensorFlow.
ANNOTATION TYPES:
- Bounding boxes for standard object detection
- Polygon labels for irregular and organic shapes
- Semantic segmentation at pixel level
- Keypoint labelling for human pose estimation
- Video annotation with frame by frame tracking IDs
- Image classification and directory sorting
TOOLS: CVAT, Roboflow, LabelImg, and Labelme
WHAT YOU GET:
- Labels exported in YOLO TXT, COCO JSON, Pascal VOC XML, or CSV
- Consistent class naming and a written labelling guide
- Train, validation, and test split on request
- Your files kept private and deleted after delivery on request
Message me before ordering with your class list, a few sample images, and your labelling rules. LOOKING FORWARD TO WORKING WITH YOU AS READ THIS!!!
Technique:
Automated
Tagging type:
Text
•
Image
•
Video
FAQ
How do I send my images, and how should they be organised?
Send a Google Drive, Dropbox, or WeTransfer link with the images in one folder. Include your class list and any labelling rules, such as whether to label partly hidden or blurred objects. Zipped folders and raw video files are both fine.
How are video frames counted against the package image limit?
Each extracted frame counts as one image. Tell me your video length and the frame rate you need, for example one frame per second, and I will work out the total count and quote the right package before you order.
What if some labels are wrong or need adjusting?
Revisions are included. I label a small sample first so the rules are agreed early, which prevents most corrections. If a class is mislabelled after delivery, send the image IDs and I will fix and re-export the dataset.
Can you handle private or proprietary images?
Yes. Your files are used only for your project, are not shared or reused, and are deleted after delivery on request. I am happy to sign an NDA through Fiverr before you upload anything sensitive.
Can you also train the model after the data is labelled?
Yes. I have a separate gig for YOLO training, detection, tracking, and counting. Message me and I can quote the labelling and the training together, so the dataset is exported exactly the way the training needs it.
