reduced-3dgs 1.8.18__tar.gz → 1.9.0__tar.gz

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: reduced_3dgs
-Version: 1.8.18
+Version: 1.9.0
 Summary: Refactored code for the paper "Reducing the Memory Footprint of 3D Gaussian Splatting"
 Home-page: https://github.com/yindaheng98/reduced-3dgs
 Author: yindaheng98

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs/combinations.py

@@ -1,13 +1,67 @@
+from typing import List
 from gaussian_splatting import GaussianModel, CameraTrainableGaussianModel, Camera
 from gaussian_splatting.dataset import CameraDataset, TrainableCameraDataset
 from gaussian_splatting.trainer import OpacityResetDensificationTrainer
 # from gaussian_splatting.trainer import BaseOpacityResetDensificationTrainer as OpacityResetDensificationTrainer
-from gaussian_splatting.trainer import OpacityResetTrainerWrapper, CameraTrainerWrapper
+from gaussian_splatting.trainer import OpacityResetTrainerWrapper, CameraTrainerWrapper, NoopDensifier, DepthTrainerWrapper
 from .shculling import VariableSHGaussianModel, SHCullingTrainerWrapper
 from .shculling import SHCullingTrainer
 # from .shculling import BaseSHCullingTrainer as SHCullingTrainer
-from .pruning import PruningTrainer, PrunerInDensifyTrainer
+from .pruning import PruningTrainerWrapper, PrunerInDensifyTrainerWrapper
 # from .pruning import BasePruningTrainer as PruningTrainer, BasePrunerInDensifyTrainer as PrunerInDensifyTrainer
+from .importance import ImportancePruner
+
+
+def BaseFullPruningTrainer(
+        model: GaussianModel,
+        scene_extent: float,
+        dataset: List[Camera],
+        *args,
+        importance_prune_from_iter=1000,
+        importance_prune_until_iter=15000,
+        importance_prune_interval: int = 100,
+        **kwargs):
+    return PruningTrainerWrapper(
+        lambda model, scene_extent, dataset: ImportancePruner(
+            NoopDensifier(model),
+            dataset,
+            importance_prune_from_iter=importance_prune_from_iter,
+            importance_prune_until_iter=importance_prune_until_iter,
+            importance_prune_interval=importance_prune_interval,
+        ),
+        model, scene_extent, dataset,
+        *args, **kwargs
+    )
+
+
+def BaseFullPrunerInDensifyTrainer(
+        model: GaussianModel,
+        scene_extent: float,
+        dataset: List[Camera],
+        *args,
+        importance_prune_from_iter=1000,
+        importance_prune_until_iter=15000,
+        importance_prune_interval: int = 100,
+        **kwargs):
+    return PrunerInDensifyTrainerWrapper(
+        lambda model, scene_extent, dataset: ImportancePruner(
+            NoopDensifier(model),
+            dataset,
+            importance_prune_from_iter=importance_prune_from_iter,
+            importance_prune_until_iter=importance_prune_until_iter,
+            importance_prune_interval=importance_prune_interval,
+        ),
+        model, scene_extent, dataset,
+        *args, **kwargs
+    )
+
+
+def DepthFullPruningTrainer(model: GaussianModel, scene_extent: float, dataset: TrainableCameraDataset, *args, **kwargs):
+    return DepthTrainerWrapper(BaseFullPruningTrainer, model, scene_extent, *args, dataset=dataset, **kwargs)
+
+
+def DepthFullPrunerInDensifyTrainer(model: GaussianModel, scene_extent: float, dataset: TrainableCameraDataset, *args, **kwargs):
+    return DepthTrainerWrapper(BaseFullPrunerInDensifyTrainer, model, scene_extent, *args, dataset=dataset, **kwargs)
 
 
 def OpacityResetPruningTrainer(
@@ -16,7 +70,7 @@ def OpacityResetPruningTrainer(
         dataset: CameraDataset,
         *args, **kwargs):
     return OpacityResetTrainerWrapper(
-        lambda model, scene_extent, *args, **kwargs: PruningTrainer(model, scene_extent, dataset, *args, **kwargs),
+        lambda model, scene_extent, *args, **kwargs: DepthFullPruningTrainer(model, scene_extent, dataset, *args, **kwargs),
         model, scene_extent,
         *args, **kwargs
     )
@@ -28,13 +82,17 @@ def OpacityResetPrunerInDensifyTrainer(
         dataset: CameraDataset,
         *args, **kwargs):
     return OpacityResetTrainerWrapper(
-        lambda model, scene_extent, *args, **kwargs: PrunerInDensifyTrainer(model, scene_extent, dataset, *args, **kwargs),
+        lambda model, scene_extent, *args, **kwargs: DepthFullPrunerInDensifyTrainer(model, scene_extent, dataset, *args, **kwargs),
         model, scene_extent,
         *args, **kwargs
     )
 
 
-def SHCullingDensifyTrainer(
+PruningTrainer = OpacityResetPruningTrainer
+PrunerInDensifyTrainer = OpacityResetPrunerInDensifyTrainer
+
+
+def SHCullingDensificationTrainer(
     model: VariableSHGaussianModel,
         scene_extent: float,
         dataset: CameraDataset,
@@ -46,7 +104,7 @@ def SHCullingDensifyTrainer(
     )
 
 
-def SHCullingPruneTrainer(
+def SHCullingPruningTrainer(
     model: VariableSHGaussianModel,
         scene_extent: float,
         dataset: CameraDataset,
@@ -58,7 +116,7 @@ def SHCullingPruneTrainer(
     )
 
 
-def SHCullingPruningDensifyTrainer(
+def SHCullingPrunerInDensifyTrainer(
     model: VariableSHGaussianModel,
         scene_extent: float,
         dataset: CameraDataset,
@@ -99,7 +157,7 @@ def CameraPruningTrainer(
     )
 
 
-def CameraOpacityResetPrunerInDensifyTrainer(
+def CameraPrunerInDensifyTrainer(
         model: CameraTrainableVariableSHGaussianModel,
         scene_extent: float,
         dataset: TrainableCameraDataset,
@@ -117,19 +175,19 @@ def CameraSHCullingDensifyTrainer(
         dataset: TrainableCameraDataset,
         *args, **kwargs):
     return CameraTrainerWrapper(
-        SHCullingDensifyTrainer,
+        SHCullingDensificationTrainer,
         model, scene_extent, dataset,
         *args, **kwargs
     )
 
 
-def CameraSHCullingPruneTrainer(
+def CameraSHCullingPruningTrainer(
         model: CameraTrainableVariableSHGaussianModel,
         scene_extent: float,
         dataset: TrainableCameraDataset,
         *args, **kwargs):
     return CameraTrainerWrapper(
-        SHCullingPruneTrainer,
+        SHCullingPruningTrainer,
         model, scene_extent, dataset,
         *args, **kwargs
     )
@@ -141,7 +199,7 @@ def CameraSHCullingPruningDensifyTrainer(
         dataset: TrainableCameraDataset,
         *args, **kwargs):
     return CameraTrainerWrapper(
-        SHCullingPruningDensifyTrainer,
+        SHCullingPrunerInDensifyTrainer,
         model, scene_extent, dataset,
         *args, **kwargs
     )

reduced_3dgs-1.9.0/reduced_3dgs/importance/__init__.py

@@ -0,0 +1,3 @@
+from .trainer import ImportancePruner, BaseImportancePruningTrainer
+from .combinations import BaseImportancePrunerInDensifyTrainer, DepthImportancePruningTrainer, DepthImportancePrunerInDensifyTrainer
+from .combinations import ImportancePruningTrainer, ImportancePrunerInDensifyTrainer

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs/importance/combinations.py

@@ -9,10 +9,11 @@ def BaseImportancePrunerInDensifyTrainer(
         model: GaussianModel,
         scene_extent: float,
         dataset: List[Camera],
+        *args,
         importance_prune_from_iter=1000,
         importance_prune_until_iter=15000,
         importance_prune_interval=100,
-        *args, **kwargs):
+        **kwargs):
     return DensificationTrainerWrapper(
         lambda model, scene_extent: ImportancePruner(
             NoopDensifier(model),

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs/importance/trainer.py

@@ -75,7 +75,7 @@ def count_render(self: GaussianModel, viewpoint_camera: Camera):
     }
 
 
-def prune_gaussians(model: GaussianModel, dataset: CameraDataset):
+def prune_list(model: GaussianModel, dataset: CameraDataset):
     gaussian_count = torch.zeros(model.get_xyz.shape[0], device=model.get_xyz.device, dtype=torch.int)
     opacity_important_score = torch.zeros(model.get_xyz.shape[0], device=model.get_xyz.device, dtype=torch.float)
     T_alpha_important_score = torch.zeros(model.get_xyz.shape[0], device=model.get_xyz.device, dtype=torch.float)
@@ -84,16 +84,66 @@ def prune_gaussians(model: GaussianModel, dataset: CameraDataset):
         gaussian_count += out["gaussians_count"]
         opacity_important_score += out["opacity_important_score"]
         T_alpha_important_score += out["T_alpha_important_score"]
-    return None
+    return gaussian_count, opacity_important_score, T_alpha_important_score
+
+
+# return importance score with adaptive volume measure described in paper
+def calculate_v_imp_score(gaussians: GaussianModel, imp_list, v_pow):
+    """
+    :param gaussians: A data structure containing Gaussian components with a get_scaling method.
+    :param imp_list: The importance scores for each Gaussian component.
+    :param v_pow: The power to which the volume ratios are raised.
+    :return: A list of adjusted values (v_list) used for pruning.
+    """
+    # Calculate the volume of each Gaussian component
+    volume = torch.prod(gaussians.get_scaling, dim=1)
+    # Determine the kth_percent_largest value
+    index = int(len(volume) * 0.9)
+    sorted_volume, _ = torch.sort(volume, descending=True)
+    kth_percent_largest = sorted_volume[index]
+    # Calculate v_list
+    v_list = torch.pow(volume / kth_percent_largest, v_pow)
+    v_list = v_list * imp_list
+    return v_list
+
+
+def score2mask(percent, import_score: list, threshold=None):
+    sorted_tensor, _ = torch.sort(import_score, dim=0)
+    index_nth_percentile = int(percent * (sorted_tensor.shape[0] - 1))
+    value_nth_percentile = sorted_tensor[index_nth_percentile]
+    thr = min(threshold, value_nth_percentile) if threshold is not None else value_nth_percentile
+    prune_mask = (import_score <= thr)
+    return prune_mask
+
+
+def prune_gaussians(gaussians: GaussianModel, dataset: CameraDataset, prune_type="important_score", prune_percent=0.1, prune_thr=None, v_pow=0.1):
+    gaussian_list, opacity_imp_list, T_alpha_imp_list = prune_list(gaussians, dataset)
+    match prune_type:
+        case "important_score":
+            mask = score2mask(prune_percent, opacity_imp_list, prune_thr)
+        case "v_important_score":
+            v_list = calculate_v_imp_score(gaussians, opacity_imp_list, v_pow)
+            mask = score2mask(prune_percent, v_list, prune_thr)
+        case "max_v_important_score":
+            v_list = opacity_imp_list * torch.max(gaussians.get_scaling, dim=1)[0]
+            mask = score2mask(prune_percent, v_list, prune_thr)
+        case "count":
+            mask = score2mask(prune_percent, gaussian_list, prune_thr)
+        case "T_alpha":
+            # new importance score defined by doji
+            mask = score2mask(prune_percent, T_alpha_imp_list, prune_thr)
+        case _:
+            raise Exception("Unsupportive prunning method")
+    return mask
 
 
 class ImportancePruner(DensifierWrapper):
     def __init__(
             self, base_densifier: AbstractDensifier,
             dataset: CameraDataset,
-            importance_prune_from_iter=1000,
-            importance_prune_until_iter=15000,
-            importance_prune_interval: int = 100,
+            importance_prune_from_iter=15000,
+            importance_prune_until_iter=20000,
+            importance_prune_interval: int = 1000,
     ):
         super().__init__(base_densifier)
         self.dataset = dataset
@@ -109,32 +159,15 @@ class ImportancePruner(DensifierWrapper):
         return ret
 
 
-def ImportancePrunerWrapper(
-        base_densifier_constructor: Callable[..., AbstractDensifier],
-        model: GaussianModel,
-        scene_extent: float,
-        dataset: List[Camera],
-        importance_prune_from_iter=1000,
-        importance_prune_until_iter=15000,
-        importance_prune_interval: int = 100,
-        *args, **kwargs):
-    return ImportancePruner(
-        base_densifier_constructor(model, scene_extent, *args, **kwargs),
-        dataset,
-        importance_prune_from_iter=importance_prune_from_iter,
-        importance_prune_until_iter=importance_prune_until_iter,
-        importance_prune_interval=importance_prune_interval,
-    )
-
-
 def BaseImportancePruningTrainer(
         model: GaussianModel,
         scene_extent: float,
         dataset: List[Camera],
+        *args,
         importance_prune_from_iter=1000,
         importance_prune_until_iter=15000,
         importance_prune_interval: int = 100,
-        *args, **kwargs):
+        **kwargs):
     return DensificationTrainer(
         model, scene_extent,
         ImportancePruner(

reduced_3dgs-1.9.0/reduced_3dgs/pruning/__init__.py

@@ -0,0 +1,2 @@
+from .trainer import BasePruner, BasePruningTrainer, PruningTrainerWrapper
+from .combinations import BasePrunerInDensifyTrainer, PruningTrainer, PrunerInDensifyTrainer, PrunerInDensifyTrainerWrapper

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs/pruning/combinations.py

@@ -1,15 +1,17 @@
 
-from typing import List
+from typing import Callable, List
 from gaussian_splatting import Camera, GaussianModel
 from gaussian_splatting.dataset import TrainableCameraDataset
-from gaussian_splatting.trainer import DepthTrainerWrapper, NoopDensifier, DensificationTrainerWrapper
+from gaussian_splatting.trainer import AbstractDensifier, DepthTrainerWrapper, NoopDensifier, SplitCloneDensifierTrainerWrapper
 from .trainer import BasePruner, BasePruningTrainer
 
 
-def BasePrunerInDensifyTrainer(
+def PrunerInDensifyTrainerWrapper(
+        noargs_base_densifier_constructor: Callable[[GaussianModel, float, List[Camera]], AbstractDensifier],
         model: GaussianModel,
         scene_extent: float,
         dataset: List[Camera],
+        *args,
         prune_from_iter=1000,
         prune_until_iter=15000,
         prune_interval: int = 100,
@@ -17,10 +19,10 @@ def BasePrunerInDensifyTrainer(
         lambda_mercy=1.,
         mercy_minimum=3,
         mercy_type='redundancy_opacity',
-        *args, **kwargs):
-    return DensificationTrainerWrapper(
+        **kwargs):
+    return SplitCloneDensifierTrainerWrapper(
         lambda model, scene_extent: BasePruner(
-            NoopDensifier(model),
+            noargs_base_densifier_constructor(model, scene_extent, dataset),
             dataset,
             prune_from_iter=prune_from_iter,
             prune_until_iter=prune_until_iter,
@@ -36,6 +38,18 @@ def BasePrunerInDensifyTrainer(
     )
 
 
+def BasePrunerInDensifyTrainer(
+        model: GaussianModel,
+        scene_extent: float,
+        dataset: List[Camera],
+        *args, **kwargs):
+    return PrunerInDensifyTrainerWrapper(
+        lambda model, scene_extent, dataset: NoopDensifier(model),
+        model, scene_extent, dataset,
+        *args, **kwargs
+    )
+
+
 # Depth trainer
 
 

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs/pruning/trainer.py

@@ -1,4 +1,4 @@
-from typing import List
+from typing import Callable, List
 import torch
 from gaussian_splatting import GaussianModel, Camera
 from gaussian_splatting.trainer import AbstractDensifier, DensifierWrapper, DensificationTrainer, NoopDensifier
@@ -108,10 +108,12 @@ class BasePruner(DensifierWrapper):
         return ret
 
 
-def BasePruningTrainer(
+def PruningTrainerWrapper(
+        noargs_base_densifier_constructor: Callable[[GaussianModel, float, List[Camera]], AbstractDensifier],
         model: GaussianModel,
         scene_extent: float,
         dataset: List[Camera],
+        *args,
         prune_from_iter=1000,
         prune_until_iter=15000,
         prune_interval: int = 100,
@@ -119,11 +121,11 @@ def BasePruningTrainer(
         lambda_mercy=1.,
         mercy_minimum=3,
         mercy_type='redundancy_opacity',
-        *args, **kwargs):
+        **kwargs):
     return DensificationTrainer(
         model, scene_extent,
         BasePruner(
-            NoopDensifier(model),
+            noargs_base_densifier_constructor(model, scene_extent, dataset),
             dataset,
             prune_from_iter=prune_from_iter,
             prune_until_iter=prune_until_iter,
@@ -134,3 +136,15 @@ def BasePruningTrainer(
             mercy_type=mercy_type,
         ), *args, **kwargs
     )
+
+
+def BasePruningTrainer(
+        model: GaussianModel,
+        scene_extent: float,
+        dataset: List[Camera],
+        *args, **kwargs):
+    return DensificationTrainer(
+        lambda model, scene_extent, dataset: NoopDensifier(model),
+        model, scene_extent, dataset,
+        *args, **kwargs
+    )

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs/train.py

@@ -14,30 +14,27 @@ from gaussian_splatting.trainer.extensions import ScaleRegularizeTrainerWrapper
 from reduced_3dgs.quantization import AbstractQuantizer, VectorQuantizeTrainerWrapper
 from reduced_3dgs.shculling import VariableSHGaussianModel, SHCullingTrainer
 from reduced_3dgs.pruning import PruningTrainer
-from reduced_3dgs.pruning.importance import ImportancePruningTrainerWrapper
-from reduced_3dgs.combinations import OpacityResetPrunerInDensifyTrainer, SHCullingDensifyTrainer, SHCullingPruneTrainer, SHCullingPruningDensifyTrainer
+from reduced_3dgs.combinations import PrunerInDensifyTrainer, SHCullingDensificationTrainer, SHCullingPruningTrainer, SHCullingPrunerInDensifyTrainer
 from reduced_3dgs.combinations import CameraTrainableVariableSHGaussianModel, CameraSHCullingTrainer, CameraPruningTrainer
-from reduced_3dgs.combinations import CameraOpacityResetPrunerInDensifyTrainer, CameraSHCullingDensifyTrainer, CameraSHCullingPruneTrainer, CameraSHCullingPruningDensifyTrainer
+from reduced_3dgs.combinations import CameraPrunerInDensifyTrainer, CameraSHCullingDensifyTrainer, CameraSHCullingPruningTrainer, CameraSHCullingPruningDensifyTrainer
 
 
 basemodes = {
     "shculling": SHCullingTrainer,
     "pruning": PruningTrainer,
-    "densify-pruning": OpacityResetPrunerInDensifyTrainer,
-    "densify-shculling": SHCullingDensifyTrainer,
-    "prune-shculling": SHCullingPruneTrainer,
-    "densify-prune-shculling": SHCullingPruningDensifyTrainer,
+    "densify-pruning": PrunerInDensifyTrainer,
+    "densify-shculling": SHCullingDensificationTrainer,
+    "prune-shculling": SHCullingPruningTrainer,
+    "densify-prune-shculling": SHCullingPrunerInDensifyTrainer,
 }
-basemodes = {k: lambda *args, **kwargs: ImportancePruningTrainerWrapper(v, *args, **kwargs) for k, v in basemodes.items()}
 cameramodes = {
     "camera-shculling": CameraSHCullingTrainer,
     "camera-pruning": CameraPruningTrainer,
-    "camera-densify-pruning": CameraOpacityResetPrunerInDensifyTrainer,
+    "camera-densify-pruning": CameraPrunerInDensifyTrainer,
     "camera-densify-shculling": CameraSHCullingDensifyTrainer,
-    "camera-prune-shculling": CameraSHCullingPruneTrainer,
+    "camera-prune-shculling": CameraSHCullingPruningTrainer,
     "camera-densify-prune-shculling": CameraSHCullingPruningDensifyTrainer,
 }
-cameramodes = {k: lambda *args, **kwargs: ImportancePruningTrainerWrapper(v, *args, **kwargs) for k, v in cameramodes.items()}
 
 
 def prepare_quantizer(

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/reduced_3dgs.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: reduced_3dgs
-Version: 1.8.18
+Version: 1.9.0
 Summary: Refactored code for the paper "Reducing the Memory Footprint of 3D Gaussian Splatting"
 Home-page: https://github.com/yindaheng98/reduced-3dgs
 Author: yindaheng98

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/setup.py

@@ -60,7 +60,7 @@ if os.name == 'nt':
 
 setup(
     name="reduced_3dgs",
-    version='1.8.18',
+    version='1.9.0',
     author='yindaheng98',
     author_email='yindaheng98@gmail.com',
     url='https://github.com/yindaheng98/reduced-3dgs',

{reduced_3dgs-1.8.18 → reduced_3dgs-1.9.0}/submodules/gaussian-importance/cuda_rasterizer/forward.cu

@@ -471,9 +471,9 @@ renderCUDA_count(
 			}
 
 			//add count 
-			gaussian_count[collected_id[j]]++;
-			opacity_important_score[collected_id[j]] += con_o.w; //opacity
-			T_alpha_important_score[collected_id[j]] += alpha * T;
+			atomicAdd(&gaussian_count[collected_id[j]], 1);
+			atomicAdd(&opacity_important_score[collected_id[j]], con_o.w); //opacity
+			atomicAdd(&T_alpha_important_score[collected_id[j]], alpha * T);
 
 			// Eq. (3) from 3D Gaussian splatting paper.
 			for (int ch = 0; ch < CHANNELS; ch++)

reduced_3dgs-1.8.18/reduced_3dgs/importance/__init__.py

@@ -1,2 +0,0 @@
-from .trainer import ImportancePruner, BaseImportancePruningTrainer, ImportancePrunerWrapper
-from .combinations import BaseImportancePrunerInDensifyTrainer, ImportancePruningTrainer, ImportancePrunerInDensifyTrainer

reduced_3dgs-1.8.18/reduced_3dgs/pruning/__init__.py

@@ -1,2 +0,0 @@
-from .trainer import BasePruner, BasePruningTrainer
-from .combinations import BasePrunerInDensifyTrainer, PruningTrainer, PrunerInDensifyTrainer