Computer Vision Engineer for Barcode & Object Detection
Skills Required
Description
This project focuses on building a robust computer vision pipeline for barcode and object detection. The engineer will need to combine accuracy, speed, and reliability in order to deliver a system ready for real-world deployment.
Python and OpenCV will be the foundation of the work. Preprocessing steps such as image cleaning, resizing, and augmentation should be applied to ensure consistent detection across varied inputs.
PyTorch will be used for training and fine-tuning deep learning models. Transfer learning from existing architectures may help speed up development while maintaining high accuracy.
The engineer will also work on barcode detection specifically. These use cases require handling different orientations, lighting conditions, and even partially damaged labels, so robust preprocessing techniques are vital.
Object detection tasks will go beyond barcodes, requiring the ability to distinguish multiple product types or items in a single frame. Models should be optimized for both precision and recall to reduce false positives and missed detections.
Data annotation will be part of the workflow. Clean, labeled datasets are crucial, and the engineer may need to assist with annotation tools or pipeline creation.
Model optimization will include reducing inference time, pruning unnecessary weights, and potentially deploying lightweight models for edge devices.
MLOps practices will be expected. Reproducibility, ver...