Talent.com
This job offer is not available in your country.
Video Annotator

Video Annotator

Innodata Inc.panchkula, haryana, in
1 day ago
Job description

Skeleton Keypoints & Bounding Box Video Annotation (Immediate joiners only)

Core Annotation Skills

  • Accurate placement of skeleton keypoints (COCO or custom schema).
  • Consistent joint alignment across frames, even with motion blur or occlusion.
  • Precise and smooth bounding boxes across all frames.
  • Skilled in handling scale, perspective, and partially visible subjects.

Video Annotation Expertise

  • Proficient in frame-by-frame labeling and interpolation.
  • Maintains temporal consistency in skeletons and bounding boxes.
  • Detects and fixes annotation drift in long sequences.
  • Balances speed and accuracy in complex datasets.

    • Cognitive & Analytical Abilities

  • Excellent attention to detail and spatial awareness.
  • Strong understanding of body posture and motion.
  • Effective in resolving ambiguous or unclear frames.
  • Ensures consistent application of labeling rules.

    • Quality Awareness

  • Familiar with IoU (bounding boxes) and PCK (keypoints) metrics.
  • Reviews and corrects annotations before submission.
  • Understands how data quality affects AI / ML model performance.

    • Preferred Background (Optional)

  • Knowledge of basic anatomy (human / animal).
  • Experience with pose estimation datasets (COCO, MPII, etc.).
  • Worked on sports, gesture, or action-tracking projects.
  • Familiarity with platforms like V7 Darwin.
    • Create a job alert for this search

    Video • panchkula, haryana, in