AI Vision System - YOLO Dataset Preparation
Budget: $250 – $750 USD
Title: Image Annotation Specialist (LabelMe) + YOLO Dataset Preparation (34 to 36 Classes)
Description:
I am building an AI-based visual inspection system and need help labeling images and preparing a dataset for YOLO training.
This is a structured, detail-oriented task. I am looking for someone reliable who can follow instructions carefully and maintain consistency across labels.
Scope of Work:
1. Image Annotation
* Use LabelMe to annotate images
* Draw bounding boxes or polygons (depending on object type)
* Label objects according to a predefined class list (36 classes total)
* Follow strict naming and labeling conventions
2. Class System
* There are **34 to 36 defect classes**
* Each class will be provided with **image samples/examples**
* You must match labels to the correct class based on those examples
* Consistency is critical (similar defects must always use the same class)
3. Dataset Preparation for YOLO
* Convert LabelMe annotations into YOLO format
* Ensure correct class IDs and structure
* Organize dataset into:
* train/
* val/
* Verify all images have matching label files
4. Quality Control
* Ensure labels are accurate and consistent
* Avoid missing or incorrect annotations
* Perform basic validation before delivery
Requirements:
* Experience with LabelMe (or similar tools like LabelImg)
* Familiarity with YOLO dataset format
* Strong attention to detail (very important)
* Ability to follow structured instructions and class definitions
Nice to Have:
* Experience with multi-class object detection datasets
* Experience working with 20+ class labeling systems
* Basic understanding of computer vision workflows
Deliverables:
* Fully labeled dataset
* YOLO-formatted label files (.txt)
* Clean folder structure ready for training
Volume:
* Initial batch: I will provide images, images must be scrubbed first to seperate classes
* Potential for ongoing work depending on performance
To Apply:
Please include:
* Example of similar annotation work (if available)
* Confirmation that you understand YOLO format
* Your expected turnaround time
* Confirmation you can handle a 34 to 36-class labeling system
Test Task:
Send sample of similar work
Notes:
* Accuracy and consistency are more important than speed
* All class definitions and example images will be provided
* This is part of a larger AI system, so quality matters
Description:
I am building an AI-based visual inspection system and need help labeling images and preparing a dataset for YOLO training.
This is a structured, detail-oriented task. I am looking for someone reliable who can follow instructions carefully and maintain consistency across labels.
Scope of Work:
1. Image Annotation
* Use LabelMe to annotate images
* Draw bounding boxes or polygons (depending on object type)
* Label objects according to a predefined class list (36 classes total)
* Follow strict naming and labeling conventions
2. Class System
* There are **34 to 36 defect classes**
* Each class will be provided with **image samples/examples**
* You must match labels to the correct class based on those examples
* Consistency is critical (similar defects must always use the same class)
3. Dataset Preparation for YOLO
* Convert LabelMe annotations into YOLO format
* Ensure correct class IDs and structure
* Organize dataset into:
* train/
* val/
* Verify all images have matching label files
4. Quality Control
* Ensure labels are accurate and consistent
* Avoid missing or incorrect annotations
* Perform basic validation before delivery
Requirements:
* Experience with LabelMe (or similar tools like LabelImg)
* Familiarity with YOLO dataset format
* Strong attention to detail (very important)
* Ability to follow structured instructions and class definitions
Nice to Have:
* Experience with multi-class object detection datasets
* Experience working with 20+ class labeling systems
* Basic understanding of computer vision workflows
Deliverables:
* Fully labeled dataset
* YOLO-formatted label files (.txt)
* Clean folder structure ready for training
Volume:
* Initial batch: I will provide images, images must be scrubbed first to seperate classes
* Potential for ongoing work depending on performance
To Apply:
Please include:
* Example of similar annotation work (if available)
* Confirmation that you understand YOLO format
* Your expected turnaround time
* Confirmation you can handle a 34 to 36-class labeling system
Test Task:
Send sample of similar work
Notes:
* Accuracy and consistency are more important than speed
* All class definitions and example images will be provided
* This is part of a larger AI system, so quality matters