Key Responsibilities
- Annotate images and video content using object detection techniques (e.g., bounding boxes, segmentation).
- Identify and extract relevant frames from long-form video recordings (e.g., 1-hour videos).
- Label video actions and activities based on predefined categories.
- Review and validate annotations produced by the team to ensure consistency and accuracy.
Scope of Work
- Data Annotation:
- Annotate approximately 1 million images across 23 use cases.
- Draw accurate bounding boxes and assign correct class labels to each object.
- Maintain consistency and high standards in labeling.
- Target: 1000 annotated images per day per annotator.
- Quality Assurance:
- Conduct self-reviews before submission.
- Revise any flagged or substandard annotations.
- Data Management:
- Track progress using project management and data labeling platforms.
- Log any issues or ambiguities encountered during annotation tasks.
Requirements
Required Skills and Qualifications
- Proven experience in image and video annotation for object detection and action recognition.
- Ability to work independently, manage time efficiently, and meet tight deadlines.
- Basic understanding of computer vision concepts is advantageous.
- Familiarity with standard annotation tools such as CVAT, Labelbox, and Roboflow.
Originally posted on Himalayas
Apply Now