Lift4D
[SIGGRAPH Asia 2026] Lift4D: Harmonizing Single-View 3D Estimation for 4D Reconstruction In-the-Wild
파일 탐색기
최종 버전 다운로드 (.zip)- __init__.py
- __init__.py
- ddim.py
- ddpm.py
- sampling_util.py
- autoencoder.py
- __init__.py
- model.py
- openaimodel.py
- util.py
- __init__.py
- distributions.py
- __init__.py
- modules.py
- attention.py
- ema.py
- x_transformer.py
- helpers.py
- id_loss.py
- model_irse.py
- lr_scheduler.py
- sd-objaverse-finetune-c_concat-256.yaml
- util.py
- __init__.py
- cameras.py
- colmap_loader.py
- dataset_readers.py
- deform_model.py
- gaussian_model.py
- auxiliary.h
- backward.cu
- backward.h
- config.h
- forward.cu
- forward.h
- rasterizer.h
- rasterizer_impl.cu
- rasterizer_impl.h
- __init__.py
- _features.hpp
- _fixes.hpp
- _noise.hpp
- _swizzle.hpp
- _swizzle_func.hpp
- _vectorize.hpp
- compute_common.hpp
- compute_vector_relational.hpp
- func_common.inl
- func_common_simd.inl
- func_exponential.inl
- func_exponential_simd.inl
- func_geometric.inl
- func_geometric_simd.inl
- func_integer.inl
- func_integer_simd.inl
- func_matrix.inl
- func_matrix_simd.inl
- func_packing.inl
- func_packing_simd.inl
- func_trigonometric.inl
- func_trigonometric_simd.inl
- func_vector_relational.inl
- func_vector_relational_simd.inl
- glm.cpp
- qualifier.hpp
- setup.hpp
- type_float.hpp
- type_half.hpp
- type_half.inl
- type_mat2x2.hpp
- type_mat2x2.inl
- type_mat2x3.hpp
- type_mat2x3.inl
- type_mat2x4.hpp
- type_mat2x4.inl
- type_mat3x2.hpp
- type_mat3x2.inl
- type_mat3x3.hpp
- type_mat3x3.inl
- type_mat3x4.hpp
- type_mat3x4.inl
- type_mat4x2.hpp
- type_mat4x2.inl
- type_mat4x3.hpp
- type_mat4x3.inl
- type_mat4x4.hpp
- type_mat4x4.inl
- type_mat4x4_simd.inl
- type_quat.hpp
- type_quat.inl
- type_quat_simd.inl
- type_vec1.hpp
- type_vec1.inl
- type_vec2.hpp
- type_vec2.inl
- type_vec3.hpp
- type_vec3.inl
- type_vec4.hpp
- type_vec4.inl
- type_vec4_simd.inl
- _matrix_vectorize.hpp
- matrix_clip_space.hpp
- matrix_clip_space.inl
- matrix_common.hpp
- matrix_common.inl
- matrix_double2x2.hpp
- matrix_double2x2_precision.hpp
- matrix_double2x3.hpp
- matrix_double2x3_precision.hpp
- matrix_double2x4.hpp
- matrix_double2x4_precision.hpp
- matrix_double3x2.hpp
- matrix_double3x2_precision.hpp
- matrix_double3x3.hpp
- matrix_double3x3_precision.hpp
- matrix_double3x4.hpp
- matrix_double3x4_precision.hpp
- matrix_double4x2.hpp
- matrix_double4x2_precision.hpp
- matrix_double4x3.hpp
- matrix_double4x3_precision.hpp
- matrix_double4x4.hpp
- matrix_double4x4_precision.hpp
- matrix_float2x2.hpp
- matrix_float2x2_precision.hpp
- matrix_float2x3.hpp
- matrix_float2x3_precision.hpp
- matrix_float2x4.hpp
- matrix_float2x4_precision.hpp
- matrix_float3x2.hpp
- matrix_float3x2_precision.hpp
- matrix_float3x3.hpp
- matrix_float3x3_precision.hpp
- matrix_float3x4.hpp
- matrix_float3x4_precision.hpp
- matrix_float4x2.hpp
- matrix_float4x2_precision.hpp
- matrix_float4x3.hpp
- matrix_float4x3_precision.hpp
- matrix_float4x4.hpp
- matrix_float4x4_precision.hpp
- matrix_int2x2.hpp
- matrix_int2x2_sized.hpp
- matrix_int2x3.hpp
- matrix_int2x3_sized.hpp
- matrix_int2x4.hpp
- matrix_int2x4_sized.hpp
- matrix_int3x2.hpp
- matrix_int3x2_sized.hpp
- matrix_int3x3.hpp
- matrix_int3x3_sized.hpp
- matrix_int3x4.hpp
- matrix_int3x4_sized.hpp
- matrix_int4x2.hpp
- matrix_int4x2_sized.hpp
- matrix_int4x3.hpp
- matrix_int4x3_sized.hpp
- matrix_int4x4.hpp
- matrix_int4x4_sized.hpp
- matrix_integer.hpp
- matrix_integer.inl
- matrix_projection.hpp
- matrix_projection.inl
- matrix_relational.hpp
- matrix_relational.inl
- matrix_transform.hpp
- matrix_transform.inl
- matrix_uint2x2.hpp
- matrix_uint2x2_sized.hpp
- matrix_uint2x3.hpp
- matrix_uint2x3_sized.hpp
- matrix_uint2x4.hpp
- matrix_uint2x4_sized.hpp
- matrix_uint3x2.hpp
- matrix_uint3x2_sized.hpp
- matrix_uint3x3.hpp
- matrix_uint3x3_sized.hpp
- matrix_uint3x4.hpp
- matrix_uint3x4_sized.hpp
- matrix_uint4x2.hpp
- matrix_uint4x2_sized.hpp
- matrix_uint4x3.hpp
- matrix_uint4x3_sized.hpp
- matrix_uint4x4.hpp
- matrix_uint4x4_sized.hpp
- quaternion_common.hpp
- quaternion_common.inl
- quaternion_common_simd.inl
- quaternion_double.hpp
- quaternion_double_precision.hpp
- quaternion_exponential.hpp
- quaternion_exponential.inl
- quaternion_float.hpp
- quaternion_float_precision.hpp
- quaternion_geometric.hpp
- quaternion_geometric.inl
- quaternion_relational.hpp
- quaternion_relational.inl
- quaternion_transform.hpp
- quaternion_transform.inl
- quaternion_trigonometric.hpp
- quaternion_trigonometric.inl
- scalar_common.hpp
- scalar_common.inl
- scalar_constants.hpp
- scalar_constants.inl
- scalar_int_sized.hpp
- scalar_integer.hpp
- scalar_integer.inl
- scalar_packing.hpp
- scalar_packing.inl
- scalar_reciprocal.hpp
- scalar_reciprocal.inl
- scalar_relational.hpp
- scalar_relational.inl
- scalar_uint_sized.hpp
- scalar_ulp.hpp
- scalar_ulp.inl
- vector_bool1.hpp
- vector_bool1_precision.hpp
- vector_bool2.hpp
- vector_bool2_precision.hpp
- vector_bool3.hpp
- vector_bool3_precision.hpp
- vector_bool4.hpp
- vector_bool4_precision.hpp
- vector_common.hpp
- vector_common.inl
- vector_double1.hpp
- vector_double1_precision.hpp
- vector_double2.hpp
- vector_double2_precision.hpp
- vector_double3.hpp
- vector_double3_precision.hpp
- vector_double4.hpp
- vector_double4_precision.hpp
- vector_float1.hpp
- vector_float1_precision.hpp
- vector_float2.hpp
- vector_float2_precision.hpp
- vector_float3.hpp
- vector_float3_precision.hpp
- vector_float4.hpp
- vector_float4_precision.hpp
- vector_int1.hpp
- vector_int1_sized.hpp
- vector_int2.hpp
- vector_int2_sized.hpp
- vector_int3.hpp
- vector_int3_sized.hpp
- vector_int4.hpp
- vector_int4_sized.hpp
- vector_integer.hpp
- vector_integer.inl
- vector_packing.hpp
- vector_packing.inl
- vector_reciprocal.hpp
- vector_reciprocal.inl
- vector_relational.hpp
- vector_relational.inl
- vector_uint1.hpp
- vector_uint1_sized.hpp
- vector_uint2.hpp
- vector_uint2_sized.hpp
- vector_uint3.hpp
- vector_uint3_sized.hpp
- vector_uint4.hpp
- vector_uint4_sized.hpp
- vector_ulp.hpp
- vector_ulp.inl
- bitfield.hpp
- bitfield.inl
- color_space.hpp
- color_space.inl
- constants.hpp
- constants.inl
- epsilon.hpp
- epsilon.inl
- integer.hpp
- integer.inl
- matrix_access.hpp
- matrix_access.inl
- matrix_integer.hpp
- matrix_inverse.hpp
- matrix_inverse.inl
- matrix_transform.hpp
- matrix_transform.inl
- noise.hpp
- noise.inl
- packing.hpp
- packing.inl
- quaternion.hpp
- quaternion.inl
- quaternion_simd.inl
- random.hpp
- random.inl
- reciprocal.hpp
- round.hpp
- round.inl
- type_aligned.hpp
- type_precision.hpp
- type_precision.inl
- type_ptr.hpp
- type_ptr.inl
- ulp.hpp
- ulp.inl
- vec1.hpp
- associated_min_max.hpp
- associated_min_max.inl
- bit.hpp
- bit.inl
- closest_point.hpp
- closest_point.inl
- color_encoding.hpp
- color_encoding.inl
- color_space.hpp
- color_space.inl
- color_space_YCoCg.hpp
- color_space_YCoCg.inl
- common.hpp
- common.inl
- compatibility.hpp
- compatibility.inl
- component_wise.hpp
- component_wise.inl
- dual_quaternion.hpp
- dual_quaternion.inl
- easing.hpp
- easing.inl
- euler_angles.hpp
- euler_angles.inl
- extend.hpp
- extend.inl
- extended_min_max.hpp
- extended_min_max.inl
- exterior_product.hpp
- exterior_product.inl
- fast_exponential.hpp
- fast_exponential.inl
- fast_square_root.hpp
- fast_square_root.inl
- fast_trigonometry.hpp
- fast_trigonometry.inl
- float_notmalize.inl
- functions.hpp
- functions.inl
- gradient_paint.hpp
- gradient_paint.inl
- handed_coordinate_space.hpp
- handed_coordinate_space.inl
- hash.hpp
- hash.inl
- integer.hpp
- integer.inl
- intersect.hpp
- intersect.inl
- io.hpp
- io.inl
- log_base.hpp
- log_base.inl
- matrix_cross_product.hpp
- matrix_cross_product.inl
- matrix_decompose.hpp
- matrix_decompose.inl
- matrix_factorisation.hpp
- matrix_factorisation.inl
- matrix_interpolation.hpp
- matrix_interpolation.inl
- matrix_major_storage.hpp
- matrix_major_storage.inl
- matrix_operation.hpp
- matrix_operation.inl
- matrix_query.hpp
- matrix_query.inl
- matrix_transform_2d.hpp
- matrix_transform_2d.inl
- mixed_product.hpp
- mixed_product.inl
- norm.hpp
- norm.inl
- normal.hpp
- normal.inl
- normalize_dot.hpp
- normalize_dot.inl
- number_precision.hpp
- number_precision.inl
- optimum_pow.hpp
- optimum_pow.inl
- orthonormalize.hpp
- orthonormalize.inl
- pca.hpp
- pca.inl
- perpendicular.hpp
- perpendicular.inl
- polar_coordinates.hpp
- polar_coordinates.inl
- projection.hpp
- projection.inl
- quaternion.hpp
- quaternion.inl
- range.hpp
- raw_data.hpp
- raw_data.inl
- rotate_normalized_axis.hpp
- rotate_normalized_axis.inl
- rotate_vector.hpp
- rotate_vector.inl
- scalar_multiplication.hpp
- scalar_relational.hpp
- scalar_relational.inl
- spline.hpp
- spline.inl
- std_based_type.hpp
- std_based_type.inl
- string_cast.hpp
- string_cast.inl
- texture.hpp
- texture.inl
- transform.hpp
- transform.inl
- transform2.hpp
- transform2.inl
- type_aligned.hpp
- type_aligned.inl
- type_trait.hpp
- type_trait.inl
- vec_swizzle.hpp
- vector_angle.hpp
- vector_angle.inl
- vector_query.hpp
- vector_query.inl
- wrap.hpp
- wrap.inl
- common.h
- exponential.h
- geometric.h
- integer.h
- matrix.h
- neon.h
- packing.h
- platform.h
- trigonometric.h
- vector_relational.h
- CMakeLists.txt
- common.hpp
- exponential.hpp
- ext.hpp
- fwd.hpp
- geometric.hpp
- glm.hpp
- integer.hpp
- mat2x2.hpp
- mat2x3.hpp
- mat2x4.hpp
- mat3x2.hpp
- mat3x3.hpp
- mat3x4.hpp
- mat4x2.hpp
- mat4x3.hpp
- mat4x4.hpp
- matrix.hpp
- packing.hpp
- trigonometric.hpp
- vec2.hpp
- vec3.hpp
- vec4.hpp
- vector_relational.hpp
- .gitignore
- CMakeLists.txt
- copying.txt
- manual.md
- readme.md
- stbi_image_write.h
- .gitignore
- .gitmodules
- CMakeLists.txt
- ext.cpp
- LICENSE.md
- rasterize_points.cu
- rasterize_points.h
- README.md
- setup.py
- .gitkeep
- ext.cpp
- LICENSE.md
- setup.py
- simple_knn.cu
- simple_knn.h
- spatial.cu
- spatial.h
- camera_utils.py
- deform_utils.py
- general_utils.py
- graphics_utils.py
- image_utils.py
- loss_utils.py
- sh_utils.py
- stable_zero123_guidance.py
- system_utils.py
- time_utils.py
- .gitignore
- LICENSE
- README.md
- requirements.txt
- train_lift4d_scgs.py
- arch.png
- intro.png
- setup.md
- default.yml
- inference.py
- hydra
- __init__.py
- utils.py
- __init__.py
- img_and_mask_transforms.py
- img_processing.py
- pose_target.py
- preprocessor.py
- transforms_3d.py
- __init__.py
- __init__.py
- utils.py
- __init__.py
- dino.py
- embedder_fuser.py
- point_remapper.py
- pointmap.py
- __init__.py
- __init__.py
- model.py
- solver.py
- __init__.py
- model.py
- __init__.py
- base.py
- classifier_free_guidance.py
- __init__.py
- base.py
- decoder_gs.py
- decoder_mesh.py
- decoder_rf.py
- encoder.py
- __init__.py
- mm_latent.py
- mot_sparse_structure_flow.py
- sparse_structure_flow.py
- sparse_structure_vae.py
- structured_latent_flow.py
- timestep_embedder.py
- __init__.py
- full_attn.py
- modules.py
- __init__.py
- full_attn.py
- masked_sdpa.py
- modules.py
- serialized_attn.py
- windowed_attn.py
- __init__.py
- conv_spconv.py
- __init__.py
- blocks.py
- modulated.py
- __init__.py
- basic.py
- linear.py
- nonlinearity.py
- norm.py
- spatial.py
- __init__.py
- blocks.py
- modulated.py
- norm.py
- spatial.py
- utils.py
- __init__.py
- gaussian_render.py
- octree_renderer.py
- sh_utils.py
- __init__.py
- gaussian_model.py
- general_utils.py
- flexicubes.py
- tables.py
- __init__.py
- cube2mesh.py
- utils_cube.py
- __init__.py
- octree_dfs.py
- __init__.py
- strivec.py
- __init__.py
- __init__.py
- postprocessing_utils.py
- random_utils.py
- render_utils.py
- __init__.py
- __init__.py
- ff.py
- __init__.py
- io.py
- base.py
- moge.py
- pointmap.py
- __init__.py
- inference_pipeline.py
- inference_pipeline_pointmap.py
- inference_utils.py
- layout_post_optimization_utils.py
- preprocess_utils.py
- video_utils.py
- plot_scene.py
- save_scene.py
- __init__.py
- image_mesh.py
- scene_visualizer.py
- __init__.py
- __init__.py
- .gitignore
- da3_utils.py
- LICENSE
- pyproject.toml
- README.md
- requirements.dev.txt
- requirements.inference.txt
- requirements.p3d.txt
- requirements.txt
- run_inference.py
- .gitignore
- LICENSE
- lift4d_datasets.py
- README.md
- segment_video.py
# 설치 가이드
1. 코드 내려받기
git clone https://github.com/yehonathanlitman/Lift4D
깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.
cd Lift4D
방금 내려받은 프로젝트 폴더 안으로 이동합니다.
2. 공식 설치 스크립트
쉬움 추천사전 준비물
- Python 3 pip 명령어를 쓰려면 Python이 필요합니다.
pip install hatchling hatch-requirements-txt ninja editables
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
pip install ./submodules/diff-gaussian-rasterization ./submodules/simple-knn
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
pip install git+https://github.com/facebookresearch/sam3.git pycocotools
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
pip install 'huggingface-hub[cli]<1.0'
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
설치 후 새 터미널을 열고, 프로그램의 버전 확인 명령(예: --version)으로 정상 설치됐는지 확인하세요.
이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.
3. CMake
보통사전 준비물
⚠️ 이 프로젝트는 규모가 큰 저장소라, 이 방법이 실제 핵심 제품이 아니라 내부 하위 패키지를 가리키는 것일 수 있습니다. README 전체를 함께 확인해보세요.
cd lift4d_scgs/submodules/diff-gaussian-rasterization
이 프로젝트의 관련 파일이 하위 폴더 안에 있어서, 먼저 그 폴더로 이동합니다.
mkdir build && cd build
빌드 결과물을 담을 폴더를 만들고 그 안으로 이동합니다.
cmake ..
소스코드를 분석해 빌드 설정 파일을 생성합니다 (build 폴더 안에서 실행해야 함).
make
생성된 빌드 설정을 바탕으로 실제 컴파일을 진행해 실행 파일을 만듭니다.
build 폴더 안에 실행 파일이 생성됐는지 확인하고, 직접 실행해보세요 (예: ./build/앱이름).
4. Python
쉬움사전 준비물
pip install -e .
requirements.txt 등에 명시된 파이썬 라이브러리를 설치합니다.
pip install -e '.[p3d]'
requirements.txt 등에 명시된 파이썬 라이브러리를 설치합니다.
pip install hatchling hatch-requirements-txt ninja editables
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
pip install -r requirements.inference.txt --no-build-isolation
requirements.txt 등에 명시된 파이썬 라이브러리를 설치합니다.
pip install -r requirements.txt
requirements.txt 등에 명시된 파이썬 라이브러리를 설치합니다.
에러 메시지 없이 실행되고 터미널에 안내 문구가 출력되면 정상입니다.
이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.
// repository documentation
Was this content helpful?
(0 ratings)