Datasets:
| { | |
| "field_id": "F007", | |
| "field": "3d shape generation", | |
| "directions": [ | |
| { | |
| "direction_id": "G0", | |
| "name": "Guided creation of new 3D assets", | |
| "scope": "Create new 3D assets that satisfy text, semantic conditions, explicit geometric information, or subject reference images; includes both per-request optimization and reusable conditional generators across requests." | |
| }, | |
| { | |
| "direction_id": "G1", | |
| "name": "Generative modeling of 3D shape families", | |
| "scope": "Learn generative rules for producing new shapes and variations. Training evidence may consist of 3D shapes or 2D observations, and the rules may explicitly model part composition, configuration, and relationships and support part-level manipulation." | |
| }, | |
| { | |
| "direction_id": "G2", | |
| "name": "Human-specific 3D generation", | |
| "scope": "Generate complete human figures or heads and faces that meet human-specific structural requirements; encompasses bodies, poses, clothing, and joints, as well as head and face identity, geometry, appearance, and expression control." | |
| }, | |
| { | |
| "direction_id": "G3", | |
| "name": "Generative modification of existing 3D assets", | |
| "scope": "Modify an existing 3D shape through deformation, stylization, or addition of detail while preserving its original form or specified attributes." | |
| } | |
| ] | |
| } | |