Machine Vision & Perception Engineer

Familiarity with: We are seeking a highly capable Machine Vision & Perception Engineer to design, implement, and deploy real-time scene understanding systems for complex, high-stakes operational environments. This role focuses on multimodal perception under extreme conditions, including low light, occlusion, motion blur, environmental variability, and limited compute resources. You will build production-grade vision systems that interpret human actions, detect and track individuals, understand context, and operate reliably on edge devices without cloud connectivity. This is a fixed-term, full-time position (12–24 months) with strong potential for renewal based on performance and program continuation. Key Responsibilities - Design and deploy real-time computer vision pipelines for: Multi-person detection, registration, and tracking Human pose estimation and skeletal tracking Body-part segmentation under occlusion Action and scene interpretation Object detection and spatial association - Develop robust perception systems that operate in: Low-light and high dynamic range environments Outdoor conditions (sunlight, weather variability) Cluttered and partially occluded scenes - Integrate multimodal sensors including: Stereo RGB / depth cameras Thermal imaging systems IMUs and localization sensors - Fuse RGB, depth, and thermal data for improved robustness - Optimize models for edge deployment (Jetson-class or similar) - Implement latency-aware, real-time inference pipelines (≤250 ms target) - Develop dynamic ID tracking across field-of-view changes - Architect and maintain production-level ML pipelines: Dataset versioning Training/validation workflows Model benchmarking under stress conditions Quantization / pruning / TensorRT optimization - Collaborate on hardware–software integration - Conduct structured field testing and performance evaluation Required Qualifications - B.S., M.S., or Ph.D. in Computer Vision, Robotics, Electrical Engineering, Computer Science, or related field - 5+ years of industry experience (or Ph.D. + 3+ years industry) in real-time perception systems - Strong experience in: Object detection (YOLO, Faster R-CNN, etc.) Human pose estimation (HRNet, OpenPose, MediaPipe, etc.) Multi-object tracking (DeepSORT, ByteTrack, etc.) Semantic/instance segmentation - Experience with low-light or adverse visual condition modeling Depth cameras (stereo or structured light) Thermal imaging systems Sensor fusion - Strong Python and C++ proficiency - Experience deploying models on edge devices (NVIDIA Jetson, embedded GPU systems) TensorRT, ONNX, model quantization ROS2 or similar robotics middleware OpenCV, PyTorch - Experience building production-grade ML systems (not just research prototypes) Preferred Qualifications Experience interpreting human actions or complex scenes Experience with ego-centric (first-person) vision Experience in robotics perception Experience with real-time constraints and power-limited systems Familiarity with localization techniques (visual odometry, IMU fusion) Experience developing for rugged or field-deployed systems Background in safety-critical or regulated environments Published research in computer vision or robotics Experience bridging academic research into production systems What Success Looks Like Reliable multi-person tracking in dynamic scenes Robust performance across lighting extremes Edge-deployable models meeting real-time constraints Clean, maintainable, production-ready perception stack Demonstrated improvement in robustness via sensor fusion Compensation & Benefits We offer a competitive compensation package including: Competitive salary based on experience Health, dental, and vision insurance Paid time off and holidays Flexible work arrangements (role dependent) Opportunity to work on high-impact, technically challenging systems Renewal potential and long-term growth opportunity based on performance and funding continuity Employment Term This is a full-time fixed-term role (12–24 months) aligned with a funded development program. Continuation or conversion to long-term employment may be available based on performance, program needs, and future funding. Ideal Candidate Profile You are: Technically rigorous Comfortable operating with ambiguity Capable of balancing research innovation with engineering discipline Excited by real-world constraints (compute, power, lighting, weather) Able to move from model training to deployed system integration Pay: $60,000.00 - $150,000.00 per year Benefits: Dental insurance Health insurance Paid time off Stock options Vision insurance Work Location: Remote