High-Accuracy Computer Vision Engineer – Cup Base Detection

We are seeking a Computer Vision Specialist to deliver a specific, well-defined computer vision task, not a long-term team role . The ideal candidate will have experience in developing computer vision algorithms and solutions, with a focus on image processing and object recognition. You will be responsible for implementing and optimizing computer vision models, and collaborating with our engineering team to integrate these solutions into our product. If you have a passion for technology and innovation, we would love to hear from you! Scope of Work: 1- Analyze production-line images 2- Detect all visible cup bases in each image 3- Accurately count cup bases per image 4- Handle real-world challenges such as: - Lighting variations - Noise and blur - Reflections from plastic surfaces - Minor perspective changes 5- Deliver a fully automated and offline solution No specific framework or algorithm is mandated. Input: 1- RGB images from an industrial production line 2- Images may include edge cases and challenging conditions Outputs: 1. Visual Output - Annotated images showing detected cup bases - Clear total count displayed per image 2. Data Output • CSV / JSON (or equivalent) containing: - Image ID - Total detected cup count - Optional confidence or validation metrics Critical Accuracy Requirement: ✅ Minimum Accuracy: 99.5% Accuracy calculation: Accuracy = (Correct Detections / Ground Truth Count) × 100 Acceptance threshold: Maximum 1 incorrect detection per 200 detections Any result below 99.5% accuracy will be rejected Validation: Tested on previously unseen images Includes normal cases and edge cases Results must be: Fully automated Repeatable Free from manual correction Constraints: No manual interaction during runtime Must work offline No external services or internet usage Deliverables: Source code and Executable system (ready to run) Usage instructions (execution only – no algorithm disclosure required) Sample output images Formal accuracy report Ideal Freelancer: Strong experience in Computer Vision Proven background in industrial inspection systems Excellent handling of real-world noisy data Ability to deliver production-level accuracy Apply tot his job