I will label egocentric video and hand keypoints for robotics ai training

Dev Eddy

Kenya

I speak English, French

There is always a solution to a problem

I build and fix training data for computer vision and robotics AI. My work covers image and video annotation, object tracking, segmentation, pose and hand keypoints, LiDAR 3D cuboids, and robot rollou...
Read More
About this Gig

Robots learn manipulation from precise first-person video labels, and general labelers often get them wrong. We specialize in egocentric and robotics data.


SERVICES

- Hand and body keypoint annotation

- Hand-object interaction: contact, grasp type, release

- Task step segmentation with timestamps and short descriptions

- Robot rollout review: success/failure and point of failure


DELIVERABLES

- Keypoint and interaction labels in JSON, COCO, or your format

- Timestamped task step file with a written description per step

- Grasp and action definition sheet, so labels stay consistent

- Rollout summary: success rate and common failure types (if ordered)

- Quality report with any unclear clips flagged


WHY WORK WITH US

- Annotators trained on manipulation and first-person footage

- Consistent definitions across your whole dataset

- Output formatted for your training pipeline

- Free revisions within your guidelines


FREE SAMPLE

Send a sample clip and your task type. We'll label it free.


Read More
label egocentric video and hand keypoints for robotics ai training
Full Screen
label egocentric video and hand keypoints for robotics ai training
Full Screen
Full Screen
label egocentric video and hand keypoints for robotics ai traininglabel egocentric video and hand keypoints for robotics ai traininglabel egocentric video and hand keypoints for robotics ai training

Technique:

Automated

Tagging type:

Text

•

Image

•

Video