AI Data Labeling & Evaluation Specialist (Freelance)
Toloka
Image
Segmentation
Object DetectionAs an AI Data Labeling & Evaluation Specialist at Toloka–Mindrift.ai, I reviewed AI-generated outputs across image, video, audio, and text modalities using structured evaluation frameworks. I identified labeling errors and inconsistencies, provided actionable feedback, and annotated large-scale datasets for model training and improvement. My work included semantic segmentation, object detection, search relevance scoring, and audio transcription validation. • Critically assessed AI outputs to ensure clarity, quality, and usability. • Performed semantic segmentation and object detection for computer vision projects. • Validated transcription quality and contextual relevance for speech recognition models. • Contributed to evaluation workflow refinement and flagged ambiguous edge cases.