Data Annotator / Egocentric Labeler at Atlas Capture (February 2026 – Present)
Execute high-precision temporal and spatial annotations for egocentric (first-person) video data following strict technical protocols. Perform action segmentation labeling to capture human activity and interactions accurately for downstream computer vision model training. Apply rigorous attention to detail to maintain consistency with dataset quality requirements and labeling guidelines. • Action segmentation & labeling for egocentric video streams • Temporal and spatial annotation aligned to strict protocols • Identification of edge cases, behavioral anomalies, and ambiguous human-object interactions • Meeting daily service-time and data-quality targets under tight schedules