Traffic Video Vehicle Annotation
Employer not named by the sourceRemote
Frontier is not the employer and does not collect applications.
About this role
Video Services, YouTube, Video Production, Video Editing, Video Processing, Computer Vision, AI Model Development, Data Annotation · I have a library of traffic-camera footage that needs detailed vehicle annotation. Your task is to go through each clip, mark every vehicle frame-by-frame, and track its position and movement across the sequence. No pedestrian or traffic-sign data are required on this project—only vehicles.
The raw material arrives as 1080p MP4s. You may work in CVAT, Labelbox, V7, or another professional tool as long as the final export is delivered in a COCO-style JSON (or compatible CSV) with unique track IDs and time-stamped coordinates. Please keep the original video resolution and frame rate intact so I can align the labels with my downstream computer-vision pipeline.
Deliverables • Annotated data file per video (JSON or CSV) • The corresponding project files from your chosen tool • A brief log noting any frames that were skipped or visually ambiguous
Acceptance criteria • 100 % of vehicles in every frame are boxed and consistently tracked • IoU ≥ 0.8 on random spot checks • File names and time stamps match the originals exactly
This is a fully remote assignment. Let me know your preferred tool and approximate throughput (minutes of video you can label per day) so we can plan