DeltaX AI India Private Limited

DeltaX is seeking an In-Cabin AI Engineer to develop state-of-the-art Driver Monitoring Systems (DMS) and Occupant Monitoring Systems (OMS). The engineer will build, optimize, and deploy in-cabin monitoring features such as gaze tracking, head pose, seatbelt detection, hand-on-wheel, and occupant safety analytics on embedded automotive boards (e.g., TI TDA4). The role requires hands-on expertise in computer vision, deep learning, and embedded AI optimization, with strong implementation skills. Basic English communication is required. Key Responsibilities: 1. Model Development: Implement computer vision tasks: Classification, Object Detection, Keypoint Detection, Pose Estimation, Segmentation. Build algorithms for DMS/OMS: gaze estimation, drowsiness, distraction, seatbelt, airbag zones, posture recognition. Develop real-time pipelines using DMS/OMS cameras and sensor fusion where applicable. 2. Model Optimization & Deployment: Convert and optimize AI models (ONNX, TensorRT, TIDL) for low-latency inference. Implement efficient pre/post-processing for real-time video streams. Optimize models using quantization, pruning, knowledge distillation. 3. Embedded Systems Integration: Integrate ICMS features with automotive ECUs and middleware. Ensure real-time performance and power compliance. 4. Validation & Testing: Conduct in-cabin testing under varied conditions (lighting, occlusions, multiple occupants). Evaluate model accuracy, latency, and robustness. Qualifications / Skills Required: Bachelors / Masters in Computer Science, Electrical Engineering, or related field. 3+ years of experience in computer vision / AI system development (preferably Automotive). Strong proficiency in Python & C++. Hands-on experience with DMS/OMS applications (face detection, gaze tracking, head pose, body/hand landmarks). Knowledge of PyTorch, TensorFlow, ONNX Runtime. Embedded AI deployment experience: TI TDA4, NVIDIA Jetson, ARM boards. Optimization using TensorRT, TIDL, OpenVX, CUDA. Expertise in OpenCV, MediaPipe, Dlib. Strong problem-solving, debugging, and basic English communication skills. Preferred Qualifications: Experience with multi-camera synchronization and calibration. Familiarity with automotive safety standards (Euro NCAP, NHTSA ICMS guidelines). Exposure to multi-modal fusion (cameras + seat sensors / CAN bus). Experience with cloud-to-edge AI model deployment pipelines.