sketchkit.sketch2model package

Subpackages

• sketchkit.sketch2model.methods package
  • Submodules
  • sketchkit.sketch2model.methods.doubao module
    • sketch_to_3d_doubao_impl()
  • sketchkit.sketch2model.methods.sens_standalone module
    • SENSStandaloneInference
      • SENSStandaloneInference._augment_cropped_square()
      • SENSStandaloneInference._check_integrity()
      • SENSStandaloneInference._download()
      • SENSStandaloneInference._find_model_dir()
      • SENSStandaloneInference._get_grid_samples()
      • SENSStandaloneInference._load_models()
      • SENSStandaloneInference._mesh_adaptive_level()
      • SENSStandaloneInference.compute_occupancy_grid()
      • SENSStandaloneInference.encode_sketch()
      • SENSStandaloneInference.infer()
      • SENSStandaloneInference.merge_zh_step_a()
      • SENSStandaloneInference.mesh_from_occupancy()
      • SENSStandaloneInference.prepare_features()
      • SENSStandaloneInference.preprocess_sketch()
      • SENSStandaloneInference.save_mesh_obj()
    • is_standalone_available()
    • sketch_to_3d_sens_standalone()
  • sketchkit.sketch2model.methods.teddy_monstermash module
    • generate_teddy_input_images()
    • preprocess_raster_for_teddy()
    • sketch_to_3d_teddy_impl()
    • teddy_from_raster()
    • validate_teddy_inputs()
  • Module contents
    • SENSStandaloneInference
      • SENSStandaloneInference._augment_cropped_square()
      • SENSStandaloneInference._check_integrity()
      • SENSStandaloneInference._download()
      • SENSStandaloneInference._find_model_dir()
      • SENSStandaloneInference._get_grid_samples()
      • SENSStandaloneInference._load_models()
      • SENSStandaloneInference._mesh_adaptive_level()
      • SENSStandaloneInference.compute_occupancy_grid()
      • SENSStandaloneInference.encode_sketch()
      • SENSStandaloneInference.infer()
      • SENSStandaloneInference.merge_zh_step_a()
      • SENSStandaloneInference.mesh_from_occupancy()
      • SENSStandaloneInference.prepare_features()
      • SENSStandaloneInference.preprocess_sketch()
      • SENSStandaloneInference.save_mesh_obj()
    • is_standalone_available()
    • sketch_to_3d_doubao_impl()
    • sketch_to_3d_sens_standalone()
    • sketch_to_3d_teddy_impl()
    • teddy_from_raster()

Submodules

sketchkit.sketch2model.modeler module

Modeler class for sketch-to-3D model conversion.

class sketchkit.sketch2model.modeler.Modeler(method: str = 'sens', device: str = 'cpu')

Bases: object

A class for sketch-to-3D model conversion.

This class provides functionality to convert sketches to 3D model files using various methods (Teddy, SENS, Doubao-Seed3D).

method

The conversion method that is used.

Type:

str

device

The device that is used (for methods that support it).

Type:

str

_run_doubao(sketch_input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Run Doubao-Seed3D method.

_run_sens(sketch_input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Run SENS method.

_run_teddy(sketch_input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Run Teddy method.

run(input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Convert sketch to 3D model.

Parameters:

• input – Input sketch, can be Sketch object, file path, or numpy array.

• output_path – Output file path (if None, will be auto-generated).

• **kwargs – Method-specific parameters.

Returns:

Output file path of the generated 3D model.

Return type:

str

Example

>>> from sketchkit.sketch2model import Modeler
>>> modeler = Modeler(method='teddy')
>>> output = modeler.run(sketch, output_path='output.obj')

Module contents

Sketch2Model - Unified Interface

Provides a unified API for sketch-to-3D model conversion using different methods.

class sketchkit.sketch2model.Modeler(method: str = 'sens', device: str = 'cpu')

Bases: object

A class for sketch-to-3D model conversion.

This class provides functionality to convert sketches to 3D model files using various methods (Teddy, SENS, Doubao-Seed3D).

method

The conversion method that is used.

Type:

str

device

The device that is used (for methods that support it).

Type:

str

_run_doubao(sketch_input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Run Doubao-Seed3D method.

_run_sens(sketch_input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Run SENS method.

_run_teddy(sketch_input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Run Teddy method.

run(input: Sketch | str | np.ndarray, output_path: str | None = None, **kwargs) → str

Convert sketch to 3D model.

Parameters:

• input – Input sketch, can be Sketch object, file path, or numpy array.

• output_path – Output file path (if None, will be auto-generated).

• **kwargs – Method-specific parameters.

Returns:

Output file path of the generated 3D model.

Return type:

str

Example

>>> from sketchkit.sketch2model import Modeler
>>> modeler = Modeler(method='teddy')
>>> output = modeler.run(sketch, output_path='output.obj')

sketchkit.sketch2model.sketch_to_3d(sketch_input: Sketch | str | np.ndarray, method: str, output_path: str | None = None, **kwargs) → str

Unified function to convert a sketch to a 3D model using various methods.

Parameters:

• sketch_input – Input sketch, can be Sketch object, file path, or numpy array

• method – Conversion method (‘teddy’, ‘sens’, or ‘doubao-seed3d’)

• output_path – Output file path (if None, will be auto-generated)

• **kwargs – Method-specific parameters

Returns:

Path to generated 3D model file

Return type:

str

Example

>>> from sketchkit.sketch2model import sketch_to_3d
>>> output = sketch_to_3d('sketch.png', method='teddy', output_path='output.obj')