Core AI & Computer Vision Systems Engineer

Mission 2030 | Aatmnirbhar Bharat · Surat, Gujarat, India · Full Time · Hybrid

Posted

Position Title: Real-Time Computer Vision & Edge AI Engineer (Founding Engineering Team / Core LLD) Reporting Structure: High-Level AI Architect (Principal ML Scientist, Google) Domain: Sub-16ms Edge AI, 3 D Pose & Shape Estimation (SMPL-X), Tensor RT C++ Inference, Zero-Copy Systems Performance Benchmark: Hard locked 60 FPS (<16.6 ms total frame budget) on dedicated RTX hardware 1. Position Overview & Architecture We are building a proprietary, ultra-low-latency spatial computing platform centered on high-fidelity 100% 3 D Digital Twin architecture and real-time human digitization. In this role, you will serve as the Low-Level Design (LLD) Core AI Engineer, working directly alongside a Lead AI Scientist from Google. Your primary mandate is to solve complex surface occlusion and volumetric estimation challenges by building an ultra-fast C++ inference pipeline. This system must accurately regress a subject's true underlying 3 D body shape and skeletal pose directly from a live camera feed. You will deploy models that extract parametric data (SMPL-X shape/pose parameters) and bridge these joint rotations seamlessly into our Vulkan graphics engine via shared GPU memory. System Architecture: ● Hardware Camera Ingestion: (V4 L2 / GStreamer / CUDA) ↓ Raw RGB Frames (Zero CPU Copy) ● Edge AI Inference: (Tensor RT / ONNX C++ API for 3 D Pose Tracking, Kinematic Anchoring, SMPL-X Shape) ↓ 3 D Skeletal Transforms & Shape Parameters ● Zero-Copy Shared Memory: (CUDA-Vulkan Bridge feeding directly into Open Rig Logic / Meta Human Engine) 2. Key Responsibilities & Deliverables A. Real-Time 3 D Pose & Shape Estimation ● Deploy and optimize state-of-the-art 3 D human body reconstruction models (e.g., Shapy, SMP Lify-X, CLIFF) to accurately regress the user's underlying skeletal structure and body volume, effectively bypassing unpredictable surface topologies and complex environmental occlusions. ● Extract mathematically stable shape parameters (β) and pose parameters (θ) to drive the skeletal hierarchy of a high-fidelity digital avatar. B. Edge Inference Pipeline (Tensor RT) ● Translate Python-based research models into production-grade C++ inference engines using NVIDIA Tensor RT and ONNX Runtime. ● Implement INT8/FP16 quantization, layer fusion, and custom CUDA plugins to ensure the entire AI inference pass executes within a strict <10 ms budget per frame. C. Temporal Smoothing & Anti-Jitter Kinematics ● Implement highly optimized temporal filters (Kalman filters, One-Euro filters, optical flow tracking) in native C++ to eliminate all high-frequency jitter from the output joint rotations before they reach the graphics engine. ● Ensure kinematic constraints (e.g., fixed bone lengths) are strictly maintained to prevent the digital asset from stretching or warping dynamically. D. Zero-Copy Ingestion & Engine Synchronization ● Build hardware-accelerated video capture pipelines using V4 L2 or GStreamer to ingest raw camera frames directly into GPU memory. ● Bridge the output coordinate data and transformation matrices to the graphics team using POSIX shared memory and CUDA-Vulkan interop (VK_KHR_external_memory_fd), eliminating CPU staging overhead. 3. Technical Qualifications & Tech Stack ● Core Programming: Production-level Modern C++ (C++17/20), Python (strictly for model training/validation), and CUDA C/C++. ● AI & Acceleration Frameworks: NVIDIA Tensor RT, ONNX Runtime (C++ API), Py Torch. ● Computer Vision Libraries: Open CV (CUDA backend), Media Pipe C++ bindings. ● Mathematical Foundations: 3 D Kinematics, Matrix Transformations, Quaternions/Euler angles, statistical body modeling (SMPL/SMPL-X architecture). ● Systems Architecture: Low-latency memory management, multi-threading (std:: jthread, lock-free queues), SIMD vectorization. 4. Relevant Projects & Demonstrable Experience (Preferred) Candidates will be preferred if they present functional codebases, Git Hub repositories, or thesis work covering: ● Real-Time Body Fitting / Pose Estimation: Practical experience deploying 3 D human pose or shape reconstruction models on live video feeds. ● Tensor RT / C++ Deployment: Demonstrable experience stripping a Py Torch model out of Python and running it natively in C++ using Tensor RT or ONNX, ideally with custom CUDA layers or INT8 calibration. ● High-Throughput Vision Pipelines: Built a C++ video processing pipeline that aggressively minimizes latency and avoids memory garbage collection pauses. ● Kinematics & Smoothing: Applied mathematical filters to raw sensor or AI data to produce smooth, mechanically accurate 3 D rotations. 5. Compensation & Engagement Structure ● Compensation: ₹1,00,000 to ₹2,00,000/month ● Mentorship: Direct architectural guidance, algorithm review, and technical leadership from a Principal ML Scientist at Google. ● Hardware: Dedicated high-end workstation equipped with discrete NVIDIA RTX hardware. HOW TO APPLY? https://forms.gle/SRndfc Zh Dv QQGh4 Y7

Apply on Mission 2030 | Aatmnirbhar Bharat's site