MCPcopy Create free account

hub / github.com/WU-CVGL/Omni-Scene / functions

Functions190 in github.com/WU-CVGL/Omni-Scene

⨍ Functions 190◇ Types & classes 33↳ Endpoints 1
Method__init__
(self)
model/losses.py:159
Method__init__
(self)
model/utils/benchmarker.py:12
Method__init__
(self, encoder=None, gs_decoder=None, use_checkpoint=False,
model/volume/volume_gs.py:14
Method__init__
(self, attn_cfgs=None, ffn_cfgs=dict( type='FFN',
model/volume/tpvformer_layer.py:49
Method__init__
(self, normalized_shape, eps=1e-6)
model/volume/vit.py:22
Method__init__
(self, dim, num_heads=8, qkv_bias=True, us
model/volume/vit.py:174
Method__init__
( self, dim, num_heads, mlp_ratio=4.0, qkv_bias=True, drop_pat
model/volume/vit.py:256
Method__init__
Args: kernel_size (Tuple): kernel size of the projection layer. stride (Tuple): stride of the projection layer.
model/volume/vit.py:314
Method__init__
(self, img_size=1024, patch_size=8, in_chans=3,
model/volume/vit.py:348
Method__init__
(self, tpv_h: int, tpv_w: int, tpv_z: int,
model/volume/cross_view_hybrid_attention.py:19
Method__init__
(self, tpv_h=200, tpv_w=200, tpv_z=16, tpv
model/volume/tpvformer_encoder.py:16
Method__init__
(self, num_feats, h, w, z,
model/volume/positional_encoding.py:11
Method__init__
( self, embed_dims=256, num_heads=8, num_levels=4, num_points=[8, 64,
model/volume/image_cross_attention.py:214
Method__init__
( self, tpv_h, tpv_w, tpv_z, pc_range, gs_dim=14, in_dims=64, hidden_dims=128, out_dims=None,
model/volume/volume_gs_decoder.py:12
Method__init__
( self, ch, sr_ratio: int = 1, num_views: int = 4, attention_op: O
model/pixel/attention_processor.py:133
Method__init__
( self, in_channels: int, out_channels: int, dropout: float = 0.0, num
model/pixel/blocks.py:40
Method__init__
( self, in_channels: int, out_channels: int, prev_output_channel: int,
model/pixel/blocks.py:134
Method__init__
(self, down_block=None, mid_block=None, up_block=None,
model/pixel/pixel_gs.py:24
Method__init__
( self, resolution: list = [224, 400], split: str = "train", use_center: bool
data/dataloader.py:103
Method__len__
(self)
data/dataloader.py:137
Method_init_weights
(self, m)
model/volume/vit.py:417
Method_is_power_of_2
(n)
model/volume/image_cross_attention.py:244
Methodbackward
(ctx, grad_scale)
model/utils/ops.py:67
Functionbinary_cross_entropy
F.binary_cross_entropy is not numerically stable in mixed-precision training.
model/utils/ops.py:373
Functionbuild
(model_config)
builder/builder.py:5
Methodbuild_covariance_from_scaling_rotation
(scaling, scaling_modifier, rotation)
model/gaussian.py:140
Functionchunk_batch
(func: Callable, chunk_size: int, *args, **kwargs)
model/utils/ops.py:111
Methodclear_history
(self)
model/utils/benchmarker.py:39
Functioncompute_absrel
( ground_truth: Float[Tensor, "batch height width"], predicted: Float[Tensor, "batch height width"], )
tools/metrics.py:50
Functionconvert_opengl_to_blender
(camera_matrix)
model/utils/camera.py:32
Methoddevice
(self)
model/volume/volume_gs.py:39
Methoddevice
(self)
model/pixel/pixel_gs.py:121
Methoddtype
(self)
model/omni_gs.py:88
Methoddtype
(self)
model/pixel/pixel_gs.py:125
Methoddump_memory
(self, path: Path)
model/utils/benchmarker.py:30
Methodforward
(self, x)
model/gaussian.py:103
Methodforward
(self, prediction)
model/losses.py:20
Methodforward
(self, inp)
model/losses.py:91
Methodforward
(self, X)
model/losses.py:128
Methodforward
(ctx, x)
model/utils/ops.py:44
Methodforward
(ctx, input_tensor, gt_grad)
model/utils/ops.py:60
Methodforward
Forward training function.
model/volume/volume_gs.py:42
Methodforward
**kwargs contains some specific arguments of attentions. Args: query (Tensor): The input query with shape
model/volume/tpvformer_layer.py:132
Methodforward
(self, x)
model/volume/vit.py:29
Methodforward
(self, x)
model/volume/vit.py:201
Methodforward
(self, x)
model/volume/vit.py:245
Methodforward
(self, x)
model/volume/vit.py:292
Methodforward
(self, x)
model/volume/vit.py:337
Methodforward
(self, x)
model/volume/vit.py:448
Methodforward
(self, query: List[Tensor], identity: Optional[List[Tensor]] = None,
model/volume/cross_view_hybrid_attention.py:143
Methodforward
Forward function. Args: mlvl_feats (tuple[Tensor]): Features from the upstream network, each is a 5D-tensor with
model/volume/tpvformer_encoder.py:268
Methodforward
(self, bs, device, ignore_axis='z')
model/volume/positional_encoding.py:26
Methodforward
Forward Function of Detr3DCrossAtten. Args: query (Tensor): Query of Transformer with shape (bs, num_query, embed
model/volume/image_cross_attention.py:66
Methodforward
Forward Function of MultiScaleDeformAttention. Args: query (Tensor): Query of Transformer with shape ( bs, num_qu
model/volume/image_cross_attention.py:355
Methodforward
tpv_list[0]: bs, h*w, c tpv_list[1]: bs, z*h, c tpv_list[2]: bs, w*z, c
model/volume/volume_gs_decoder.py:89
Methodforward
(self, hidden_states: torch.FloatTensor, scale: float = 1.0)
model/pixel/blocks.py:118
Methodforward
( self, hidden_states: torch.FloatTensor, res_samples_tuple: tuple, temb: Optional[tor
model/pixel/blocks.py:214
Methodforward
(self, hidden_states: torch.FloatTensor, scale: float = 1.0)
model/pixel/blocks.py:302
Methodforward
Forward training function.
model/pixel/pixel_gs.py:138
Methodfrom_pretrained
(cls, name="vgg_lpips")
model/losses.py:61
Functionget_activation
(name)
model/utils/ops.py:76
Functionget_camera
(num_frames, elevation=15, azimuth_start=0, azimuth_span=360, opencv_coord=False, cam_dist=1.0)
model/utils/camera.py:70
Functionget_fov_gaussian
(P)
model/utils/ops.py:348
Functionget_full_projection_matrix
( c2w: Float[Tensor, "B 4 4"], proj_mtx: Float[Tensor, "B 4 4"] )
model/utils/ops.py:296
Functionget_mvp_matrix
( c2w: Float[Tensor, "B 4 4"], proj_mtx: Float[Tensor, "B 4 4"] )
model/utils/ops.py:282
Functionget_projection_matrix
( fovy: Float[Tensor, "B"], aspect_wh: float, near: float, far: float )
model/utils/ops.py:267
Functionget_projection_matrix_gaussian_lgm
(znear, zfar, fovX, fovY, device="cuda")
model/utils/ops.py:334
Methodinit_weights
Default initialization for Parameters of Module.
model/volume/cross_view_hybrid_attention.py:53
Methodinit_weights
Default initialization for Parameters of Module.
model/volume/image_cross_attention.py:282
Functioninterpolate_intrinsics
( initial: Float[Tensor, "*#batch 3 3"], final: Float[Tensor, "*#batch 3 3"], t: Float[Tensor, " t
model/utils/interpolation.py:8
Functionload_lidar_info
(info)
data/transforms/loading.py:112
Methodload_ply
(self, path, compatible=True)
model/gaussian.py:309
Functionload_sparse_depths
(points, lidar2sensor, w2c, ck)
data/transforms/loading.py:122
Functionnormalize_camera
normalize the camera location onto a unit-sphere
model/utils/camera.py:55
Functionpass_print
(*args, **kwargs)
evaluate.py:28
Functionpass_print
(*args, **kwargs)
demo.py:27
Functionperpendicular_component
(x: Float[Tensor, "B C H W"], y: Float[Tensor, "B C H W"])
model/utils/ops.py:402
Methodplucker_embedder
( self, rays_o, rays_d )
model/pixel/pixel_gs.py:128
Functionreflect
(x, n)
model/utils/ops.py:18
Functionrescale_intrisic
rescale camera intrinsic parameter fx, fy, cx, cy to fit in target image resolution
model/utils/camera.py:80
Functionresize_image
(input_image, resolution)
model/utils/image.py:25
Methodsave_vis
(prefix, i, save_dir, n_rand_view, render_pkg, gaussians, rgbs_gt, depths_m_gt, mask_dptm, renderer)
model/omni_gs.py:499
Functionscale_tensor
( dat: Num[Tensor, "... D"], inp_scale: ValidScale, tgt_scale: ValidScale )
model/utils/ops.py:25
Methodset_loss
(key, split, loss_value, loss_weight=1.0)
model/omni_gs.py:225
Functionshifted_cosine_decay
(a, b, c, r)
model/utils/ops.py:398
Functionshifted_expotional_decay
(a, b, c, r)
model/utils/ops.py:394
Functionsigmoid_scaling
(scaling:torch.Tensor, lower_bound=0.005, upper_bound=0.02)
model/volume/volume_gs_decoder.py:6
Methodsummarize
(self)
model/utils/benchmarker.py:35
Functiontet_sdf_diff
( vert_sdf: Float[Tensor, "Nv 1"], tet_edges: Integer[Tensor, "Ne 2"] )
model/utils/ops.py:380
Functionvalidate_empty_rays
(ray_indices, t_start, t_end)
model/utils/ops.py:415
← previous101–190 of 190, ranked by callers