File size: 10,223 Bytes
684943d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
#
# Copyright (C) 2023, Inria
# GRAPHDECO research group, https://team.inria.fr/graphdeco
# All rights reserved.
#
# This software is free for non-commercial, research and evaluation use
# under the terms of the LICENSE.md file.
#
# For inquiries contact  [email protected]
#

import math

import torch
from diff_LangSurf_rasterization import \
    GaussianRasterizationSettings as PlaneGaussianRasterizationSettings
from diff_LangSurf_rasterization import \
    GaussianRasterizer as PlaneGaussianRasterizer

from field_construction.scene.app_model import AppModel
from field_construction.scene.gaussian_model import GaussianModel
from field_construction.utils.graphics_utils import normal_from_depth_image
from field_construction.utils.pose_utils import (get_camera_from_tensor,
                                                 quadmultiply)
from field_construction.utils.sh_utils import eval_sh


def render_normal(viewpoint_cam, depth, offset=None, normal=None, scale=1):
    # depth: (H, W), bg_color: (3), alpha: (H, W)
    # normal_ref: (3, H, W)
    intrinsic_matrix, extrinsic_matrix = viewpoint_cam.get_calib_matrix_nerf(scale=scale)
    st = max(int(scale/2)-1,0)
    if offset is not None:
        offset = offset[st::scale,st::scale]
    normal_ref = normal_from_depth_image(depth[st::scale,st::scale], 
                                            intrinsic_matrix.to(depth.device), 
                                            extrinsic_matrix.to(depth.device), offset)

    normal_ref = normal_ref.permute(2,0,1)
    return normal_ref

def render(
    viewpoint_camera,
    pc : GaussianModel,
    pipe, 
    bg_color : torch.Tensor, 
    scaling_modifier=1.0, 
    override_color=None, 
    app_model: AppModel=None,
    return_plane=True,
    return_depth_normal=True, 
    include_feature=True, 
    camera_pose=None
):
    """
    Render the scene. 
    
    Background tensor (bg_color) must be on GPU!
    """
    # Create zero tensor. We will use it to make pytorch return gradients of the 2D (screen-space) means
    screenspace_points = torch.zeros_like(pc.get_xyz, dtype=pc.get_xyz.dtype, requires_grad=True, device="cuda") + 0
    screenspace_points_abs = torch.zeros_like(pc.get_xyz, dtype=pc.get_xyz.dtype, requires_grad=True, device="cuda") + 0
    try:
        screenspace_points.retain_grad()
        screenspace_points_abs.retain_grad()
    except:
        pass

    # Set up rasterization configuration
    tanfovx = math.tan(viewpoint_camera.FoVx * 0.5)
    tanfovy = math.tan(viewpoint_camera.FoVy * 0.5)

    w2c = torch.eye(4).cuda()
    projmatrix = (
        w2c.unsqueeze(0).bmm(viewpoint_camera.projection_matrix.unsqueeze(0))
    ).squeeze(0)
    camera_pos = w2c.inverse()[3, :3]
    
    if camera_pose is not None:
        rel_w2c = get_camera_from_tensor(camera_pose)
        gaussians_xyz = pc._xyz.clone()
        gaussians_rot = pc._rotation.clone()
        xyz_ones = torch.ones(gaussians_xyz.shape[0], 1).cuda().float()
        xyz_homo = torch.cat((gaussians_xyz, xyz_ones), dim=1)
        gaussians_xyz_trans = (rel_w2c @ xyz_homo.T).T[:, :3]
        gaussians_rot_trans = quadmultiply(camera_pose[:4], gaussians_rot)
        means3D = gaussians_xyz_trans 
    
    else:
        means3D = pc.get_xyz

    means2D = screenspace_points
    means2D_abs = screenspace_points_abs
    opacity = pc.get_opacity
    
    # If precomputed 3d covariance is provided, use it. If not, then it will be computed from
    # scaling / rotation by the rasterizer.
    scales = None
    rotations = None
    cov3D_precomp = None
    if pipe.compute_cov3D_python:
        cov3D_precomp = pc.get_covariance(scaling_modifier)
    else:
        scales = pc.get_scaling
        rotations = gaussians_rot_trans if camera_pose is not None else pc.get_rotation
        # rotations = pc.get_rotation
    
    # If precomputed colors are provided, use them. Otherwise, if it is desired to precompute colors
    # from SHs in Python, do it. If not, then SH -> RGB conversion will be done by rasterizer.
    shs = None
    colors_precomp = None

    if override_color is None:
        if pipe.convert_SHs_python:
            shs_view = pc.get_features.transpose(1, 2).view(-1, 3, (pc.max_sh_degree+1)**2)
            dir_pp = (pc.get_xyz - viewpoint_camera.camera_center.repeat(pc.get_features.shape[0], 1))
            dir_pp_normalized = dir_pp/dir_pp.norm(dim=1, keepdim=True)
            sh2rgb = eval_sh(pc.active_sh_degree, shs_view, dir_pp_normalized)
            colors_precomp = torch.clamp_min(sh2rgb + 0.5, 0.0)
        else:
            shs = pc.get_features
    else:
        colors_precomp = override_color

    if include_feature:
        language_feature_precomp = pc.get_language_feature
        instance_feature_precomp = pc.get_instance_feature
        # language_feature_precomp = language_feature_precomp / (language_feature_precomp.norm(dim=-1, keepdim=True) + 1e-9)
        # instance_feature_precomp = instance_feature_precomp / (instance_feature_precomp.norm(dim=-1, keepdim=True) + 1e-9)
        # language_feature_precomp = torch.sigmoid(language_feature_precomp)
    else:
        language_feature_precomp = torch.zeros((1,), dtype=opacity.dtype, device=opacity.device)
        instance_feature_precomp = torch.zeros((1,), dtype=opacity.dtype, device=opacity.device)

    return_dict = None
    raster_settings = PlaneGaussianRasterizationSettings(
            image_height=int(viewpoint_camera.image_height),
            image_width=int(viewpoint_camera.image_width),
            tanfovx=tanfovx,
            tanfovy=tanfovy,
            bg=bg_color,
            scale_modifier=scaling_modifier,
            # viewmatrix=viewpoint_camera.world_view_transform,
            # projmatrix=viewpoint_camera.full_proj_transform,
            viewmatrix=w2c if camera_pose is not None else viewpoint_camera.world_view_transform,
            projmatrix=projmatrix if camera_pose is not None else viewpoint_camera.full_proj_transform,
            sh_degree=pc.active_sh_degree,
            # campos=viewpoint_camera.camera_center,
            campos=camera_pos if camera_pose is not None else viewpoint_camera.camera_center,
            prefiltered=False,
            render_geo=return_plane,
            debug=pipe.debug,
            include_feature=include_feature,
        )

    rasterizer = PlaneGaussianRasterizer(raster_settings=raster_settings)


    if not return_plane:
        rendered_image, language_feature, instance_feature, radii, out_observe, _, _ = rasterizer(
            means3D = means3D,
            means2D = means2D,
            means2D_abs = means2D_abs,
            shs = shs,
            colors_precomp = colors_precomp,
            language_feature_precomp = language_feature_precomp,
            language_feature_instance_precomp = instance_feature_precomp,
            opacities = opacity,
            scales = scales,
            rotations = rotations,
            cov3D_precomp = cov3D_precomp)
        
        return_dict =  {"render": rendered_image,
                        "viewspace_points": screenspace_points,
                        "viewspace_points_abs": screenspace_points_abs,
                        "visibility_filter" : radii > 0,
                        "radii": radii,
                        "out_observe": out_observe,
                        "language_feature": language_feature,
                        "instance_feature": instance_feature,
                        }
        if app_model is not None and pc.use_app:
            appear_ab = app_model.appear_ab[torch.tensor(viewpoint_camera.uid).cuda()]
            app_image = torch.exp(appear_ab[0]) * rendered_image + appear_ab[1]
            return_dict.update({"app_image": app_image})
        return return_dict

    global_normal = pc.get_normal(viewpoint_camera)
    local_normal = global_normal @ viewpoint_camera.world_view_transform[:3,:3]
    pts_in_cam = means3D @ viewpoint_camera.world_view_transform[:3,:3] + viewpoint_camera.world_view_transform[3,:3]
    depth_z = pts_in_cam[:, 2]
    local_distance = (local_normal * pts_in_cam).sum(-1).abs()
    input_all_map = torch.zeros((means3D.shape[0], 5)).cuda().float()
    input_all_map[:, :3] = local_normal
    input_all_map[:, 3] = 1.0
    input_all_map[:, 4] = local_distance

    rendered_image, language_feature, instance_feature, radii, out_observe, out_all_map, plane_depth = rasterizer(
        means3D = means3D,
        means2D = means2D,
        means2D_abs = means2D_abs,
        shs = shs,
        colors_precomp = colors_precomp,
        language_feature_precomp = language_feature_precomp,
        language_feature_instance_precomp = instance_feature_precomp,
        opacities = opacity,
        scales = scales,
        rotations = rotations,
        all_map = input_all_map,
        cov3D_precomp = cov3D_precomp)

    rendered_normal = out_all_map[0:3]
    rendered_alpha = out_all_map[3:4, ]
    rendered_distance = out_all_map[4:5, ]
    
    return_dict =  {"render": rendered_image,
                    "viewspace_points": screenspace_points,
                    "viewspace_points_abs": screenspace_points_abs,
                    "visibility_filter" : radii > 0,
                    "radii": radii,
                    "out_observe": out_observe,
                    "rendered_normal": rendered_normal,
                    "plane_depth": plane_depth,
                    "rendered_distance": rendered_distance,
                    "language_feature": language_feature,
                    "instance_feature": instance_feature,
                    }
    if app_model is not None:
        appear_ab = app_model.appear_ab[torch.tensor(viewpoint_camera.uid).cuda()]
        app_image = torch.exp(appear_ab[0]) * rendered_image + appear_ab[1]
        return_dict.update({"app_image": app_image})   

    if return_depth_normal:
        depth_normal = render_normal(viewpoint_camera, plane_depth.squeeze()) * (rendered_alpha).detach()
        return_dict.update({"depth_normal": depth_normal})
    
    # Those Gaussians that were frustum culled or had a radius of 0 were not visible.
    # They will be excluded from value updates used in the splitting criteria.
    return return_dict