Senior Media Pipeline Engineer / Image Processing Expert

We are seeking a highly skilled Image Processing Engineer, Steganography Expert, or Generative AI Pipeline Engineer to develop a robust workflow for purifying print-ready visual assets. The core objective is to reliably strip resilient, multi-bit cryptographic frequency patterns (such as embedded latent noise matrices) from generated images while maintaining complete, print-grade structural integrity. The ideal candidate must deeply understand frequency-domain pixel manipulation and latent space architecture to achieve a visually indistinguishable recreation without softening textures, shifting colors, or losing resolution. Scope of Work & Objectives • Frequency-Domain Purification: Isolate and disrupt sub-visual, statistical noise signatures woven into pixel values without compromising visible macro-details or edge sharpness. • Mass Batch Processing: Build a scalable, recursive script or pipeline capable of processing entire folder structures without altering filenames, directory architectures, or downscaling native resolutions. • Metadata Sanitization: Ensure deep cleansing of modern tracking vectors, including the complete removal of C2PA manifests, EXIF data, and structural file headers. Initial Paid Trial (Sample Phase) -We will begin with a funded milestone for a single sample photo. To evaluate your technical capability, you will be provided with a high-resolution reference photograph. You must deliver: 1. The Purified Asset: A high-fidelity file (.jpg or .png) maintaining the exact original filename, resolution, and pixel-for-pixel visual mapping. 2. Visual Difference Proof (SSIM Benchmark): The output must achieve a Structural Similarity Index (SSIM) of \(\ge \) 99.9%. 3. Verification Receipts: Clear log files or scan receipts demonstrating that the asset successfully clears major identification frameworks Status: Watermark Not Detected on targeted multi-bit cryptographic scanners. Classifier Status: \(\sim \)0% AI-generated / 100% Organic Human score on industry-standard platforms. How to Apply Please briefly outline your technical approach (e.g., frequency separation filters, low-denoise generative re-synthesis, or programmatic array perturbations) and mention the specific local tools or libraries you intend to leverage for the trial phase.