AI-Driven Environment 3D Modeling
Budget / Salary$250–750
TypeFreelance project
LocationRemote
Posted2 hours ago
I need an AI-based workflow that can automatically generate 3D environment assets with a solid mid-level of detail—more than simple block-outs, but not full cinematic realism. The goal is to have the system output textured, component-rich scenes I can drop straight into my pipeline for further polish.
Here is what matters most:
• The AI must generate complete environments, not just single props.
• Medium detail is key: think coherent modular pieces, clean UVs, and believable textures without heavy photoreal requirements.
• Final assets should arrive in a common format (.fbx or .obj) with PBR texture sets.
Deliverables
1. A repeatable AI workflow or script (Python preferred) that produces the models.
2. At least two sample environments created with that workflow so I can judge quality.
3. Clear setup instructions so I can run the process locally (GPU: RTX 30-series).
Acceptance Criteria
• Geometry is free of major artifacts and exports without errors.
• Textures align correctly on meshes and use AO, Albedo, Roughness, Metalness maps.
• Each sample environment stays below 500k tris and loads into Blender and Unreal Engine without issues.
If you already have experience combining tools like Blender, Houdini, Stable Diffusion, or similar generative-AI frameworks, that will speed things up. Send a brief note on your approach and a link to any relevant work, and we can get started right away.
Here is what matters most:
• The AI must generate complete environments, not just single props.
• Medium detail is key: think coherent modular pieces, clean UVs, and believable textures without heavy photoreal requirements.
• Final assets should arrive in a common format (.fbx or .obj) with PBR texture sets.
Deliverables
1. A repeatable AI workflow or script (Python preferred) that produces the models.
2. At least two sample environments created with that workflow so I can judge quality.
3. Clear setup instructions so I can run the process locally (GPU: RTX 30-series).
Acceptance Criteria
• Geometry is free of major artifacts and exports without errors.
• Textures align correctly on meshes and use AO, Albedo, Roughness, Metalness maps.
• Each sample environment stays below 500k tris and loads into Blender and Unreal Engine without issues.
If you already have experience combining tools like Blender, Houdini, Stable Diffusion, or similar generative-AI frameworks, that will speed things up. Send a brief note on your approach and a link to any relevant work, and we can get started right away.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.