Machine Learning Engineer - 360° Video/Image Processing

Job Description:

The production team at BMAD is looking for a highly motivated Machine Learning Engineer with expertise in 360° video and image processing to lead the development of AI tools designed to merge images captured from multiple cameras into seamless 360-degree images or videos, creating an immersive viewing experience

As a Machine Learning Engineer focused on generating high-quality equirectangular panoramas, your daily responsibilities will center around independently designing and building AI-driven tools and processing pipelines that transform images from multiple cameras into seamless 360° panoramas. This involves developing algorithms for image stitching, blending, and inpainting to ensure smooth transitions and minimize visible seams or distortions, particularly in challenging areas like the poles and edges. You will also work on optimizing projection techniques to accurately map 2D images onto a spherical surface, converting them into the standard 2:1 equirectangular format required for VR and immersive viewing experiences. The role requires expertise in handling large image datasets, applying advanced computer vision and deep learning methods, and ensuring that the final panoramas are visually coherent and high-resolution, ready for integration into VR platforms or further editing workflows

This is a hands-on, independent role suited for a passionate engineer with a strong home-lab

ethos; excellent knowledge of Linux-based containerized technologies, and a fascination with

bleeding-edge GPU hardware and GenAI systems.

Responsibilities:

Design, develop, potentially train, and deploy AI models for 360° image and video generation.
Develop pipelines to stitch images/videos from multiple cameras into seamless 360° visuals.
Optimize diffusion models using PyTorch, TensorRT, and related toolkits for maximum performance, accuracy, and scalability.
Procure, manage, and curate large 360° image/video datasets for model training.
Integrate AI solutions into VR-focused platforms and services.
Deploy and manage AI services using Docker, Kubernetes, and other containerized technologies on Linux systems.
Stay current with cutting-edge research and emerging trends in 360° imaging, spherical CNNs, and generative AI.
Collaborate with the production team and contribute as the AI subject matter expert on project direction and technical decision-making.
Requirements:

At least 4 years of experience in a development or research role focused on 360° image and video processing.
Proven experience stitching multi-camera content into seamless 360° images/videos.
Bachelor’s or Master’s degree in Computer Science, Mathematics, Engineering, or a computational science-related field.
Hands-on experience building and managing home servers, homelabs, and deep knowledge of Linux systems and GPU hardware.
Deep experience developing stable diffusion pipelines, model checkpoints, and LoRAs, particularly for 360° equirectangular panorama generation.
Proficiency in Python, with familiarity in C++, and Node.js being a plus.
Strong command of AI/ML tools and frameworks: PyTorch, Hugging Face Diffusers, and TensorRT.
Experience with Docker, Kubernetes, and containerized AI deployment on Linux systems.
Strong ability to work autonomously in an independent, self-directed environment.
Job Types: Full-time, Permanent

Schedule:

Monday to Friday
Work Location: Remote