Vision Algorithm Engineer

Location: Seoul Capital Area

Contract Type: Permanent

Salary: ₩60,000,000 – ₩120,000,000 per annum

A technology company is searching for a Vision Algorithm Engineer. The qualified candidate will develop and optimise modular and end-to-end vision algorithms using large-scale simulation platforms, deep learning, and reinforcement learning.

Responsibilities:

• Build and optimise large-scale simulation-based learning and evaluation pipelines
• Design and maintain high-fidelity simulation platforms for SIL, HIL, and VIL testing
• Enhance physics-based sensor models and generate realistic synthetic data for robustness testing
• Implement agent modelling with RL or data-driven behaviour models to evaluate responsiveness and social compliance
• Define KPIs and automate feedback loops for continuous algorithm improvement
• Lead development of closed-loop learning pipelines for training E2E driving policies
• Apply Sim2Real transfer strategies and develop data pipelines for continuous model enhancement
• Generate edge cases and adversarial scenarios to strengthen model robustness
• Collaborate with cross-functional teams and provide mentorship to junior engineers

Requirements:

• Master’s or Ph.D. in Computer Science, Mechanical, Electrical, Aerospace Engineering, Robotics, or a related field
• More than 7 years of experience in simulation, testing, and validation of complex systems like ADAS/autonomous driving or robotics
• Expert-level proficiency in Python and strong C++ skills for building test automation frameworks and integrating with simulation environments
• Hands-on experience with one or more industry-standard autonomous driving simulators such as CARLA, Vires VTD, NVIDIA Drive Sim, or LGSVL, and their underlying game engines (Unreal/Unity)
• Preferred requirements:
  • Experience with cloud computing platforms (AWS, Azure, GCP) and container
    technologies (Docker, Kubernetes) for scalable simulation
  • Experience with HIL test benches and real-time systems
  • Knowledge of safety standards such as ISO 26262, SOTIF (ISO 21448), and UL 4600
  • Background in statistics, particularly for rare-event simulation (Monte Carlo methods, importance sampling)

About the Company:

A leading provider of AI-based vision processing technology for autonomous vehicles and advanced driver assistance systems (ADAS). The company develops deep learning–powered software that enables vehicles to detect and recognise objects with high accuracy and efficiency.