Road Scene Image Segmentation for Autonomous Driving
Image
SegmentationI worked on a road scene image segmentation project for an autonomous driving dataset. Scope: Labeled diverse urban and highway images captured from vehicle cameras. Performed pixel-level segmentation for classes including roads, lane markings, vehicles, pedestrians, buildings, sidewalks, traffic signs, poles, vegetation, and sky. Used annotation tools such as Label Studio and CVAT for image segmentation tasks. Followed project guidelines and class taxonomy to ensure annotation consistency and high-quality outputs. Project Size: Annotated and reviewed over 12,000 images with multiple object classes per image. Worked on large-scale datasets under strict quality and deadline requirements. Quality Measures: Maintained high annotation accuracy by following detailed labeling instructions. Participated in quality reviews and corrected inconsistencies based on feedback. Achieved consistent performance scores above project quality benchmarks. I am detail-oriented, able to work independently, and capable of handling repetitive tasks while maintaining accuracy and efficiency. Additional Information Familiar with AI data annotation workflows and computer vision datasets. Comfortable using annotation platforms including CVAT, Label Studio, and MakeSense AI. Strong attention to detail and ability to meet project deadlines in remote work environments. Good communication and teamwork skills when collaborating with project reviewers and QA teams.