Computer Vision Engineer

- Image Recognition and Processing • Research and develop object recognition algorithms based on computer vision, ensuring high-precision identification and real-time processing capabilities. • Handle complex scenarios such as varying lighting conditions, angles, and occlusions to optimize recognition stability. - Deep Learning Model Training and Optimization • Design, train, and optimize deep learning models to improve recognition accuracy. • Utilize large-scale datasets for annotation, data augmentation, transfer learning, and model compression. • Model compression is a technique that condenses knowledge from multiple models into a single neural network. This involves first training a high-performance ensemble model, then training a smaller model on pseudo-data labeled by the ensemble. - Object Detection and Tracking • Implement precise detection and tracking using object detection algorithms such as YOLO, Faster R-CNN, and DETR. • Integrate Multi-Object Tracking (MOT) al...