MCPcopy Create free account

hub / github.com/InternRobotics/G2VLM / functions

Functions2,118 in github.com/InternRobotics/G2VLM

↓ 1 callersMethodinit_weights
(self)
eval_code/recons/models/vggt/layers/vision_transformer.py:176
↓ 1 callersMethodinit_weights
(self)
eval_code/recons/models/pi3/models/dinov2/models/vision_transformer.py:174
↓ 1 callersMethodinitialize_weights
(self)
eval_code/recons/models/vggt/heads/track_modules/blocks.py:90
↓ 1 callersFunctioninterpolate_extrinsics
Interpolate extrinsics between two camera poses. Linear interpolation for translation, spherical linear interpolation for rotation. Args:
eval_code/recons/models/moge/utils3d/torch/transforms.py:1069
↓ 1 callersMethodinterpolate_pos_encoding
This method allows to interpolate the pre-trained position encodings, to be able to use the model on higher resolution images. This i
modeling/dinov2_with_registers/modeling_dinov2_with_registers.py:93
↓ 1 callersMethodinterpolate_pos_encoding
This method allows to interpolate the pre-trained position encodings, to be able to use the model on higher resolution images. This i
modeling/dinov2_with_registers/modular_dinov2_with_registers.py:195
↓ 1 callersMethodinterpolate_pos_encoding
(self, x, w, h)
modeling/pi3/models/dinov2/models/vision_transformer.py:181
↓ 1 callersMethodinterpolate_pos_encoding
(self, x, w, h)
eval_code/recons/models/moge/model/dinov2/models/vision_transformer.py:179
↓ 1 callersMethodinterpolate_pos_encoding
(self, x, w, h)
eval_code/recons/models/vggt/layers/vision_transformer.py:183
↓ 1 callersMethodinterpolate_pos_encoding
(self, x, w, h)
eval_code/recons/models/pi3/models/dinov2/models/vision_transformer.py:181
↓ 1 callersFunctioninterpolate_transform
(transform1: torch.Tensor, transform2: torch.Tensor, t: Union[Number, torch.Tensor])
eval_code/recons/models/moge/utils3d/torch/transforms.py:1097
↓ 1 callersFunctionintrinsics_from_focal_center
Get OpenCV intrinsics matrix Returns: (np.ndarray): [..., 3, 3] OpenCV intrinsics matrix
eval_code/recons/models/moge/utils3d/numpy/transforms.py:132
↓ 1 callersFunctionis_logging_process
()
modeling/pi3/utils/basic.py:67
↓ 1 callersFunctionis_port_in_use
(host, port)
modeling/pi3/utils/debug.py:36
↓ 1 callersFunctionis_port_in_use
(host, port)
eval_code/recons/models/pi3/utils/debug.py:36
↓ 1 callersFunctionkaiming_init
(module, a=0, mode='fan_out', nonlinearity='relu',
modeling/pi3/models/segformer/head.py:56
↓ 1 callersFunctionlaplacian
Laplacian smooth with cotangent weights Args: vertices (torch.Tensor): shape (..., N, 3) faces (torch.Tensor): shape (T, 3)
eval_code/recons/models/moge/utils3d/torch/mesh.py:620
↓ 1 callersFunctionlerp_se3_matrix
Linear interpolation between two SE(3) matrices. Args: T1 (np.ndarray): [..., 4, 4] SE(3) matrix 1 T2 (np.ndarray): [..., 4,
eval_code/recons/models/moge/utils3d/numpy/transforms.py:1010
↓ 1 callersMethodlink
Links the output of the source node to the input of the destination node. If the source or destination node is None, the pipeline's i
eval_code/recons/models/moge/utils/pipeline.py:259
↓ 1 callersFunctionload_and_preprocess_images
A quick start function to load and preprocess images for model input. This assumes the images should have the same shape for easier batching,
eval_code/recons/models/vggt/utils/load_fn.py:12
↓ 1 callersFunctionload_and_resize14
(images, new_width)
data/transforms_vggt.py:454
↓ 1 callersFunctionload_cam_mvsnet
read camera txt file
eval_code/recons/datasets/dtu.py:21
↓ 1 callersFunctionload_images
Loads images from a directory or video, resizes them to a uniform size, then converts and stacks them into a single [N, 3, H, W] PyTorch tens
eval_code/recons/interfaces/g2vlm.py:21
↓ 1 callersFunctionload_images
Loads images from a directory or video, resizes them to a uniform size, then converts and stacks them into a single [N, 3, H, W] PyTorch tens
eval_code/recons/interfaces/pi3.py:18
↓ 1 callersFunctionload_model_and_tokenizer_from_path
(model_path)
g2vlm_utils.py:32
↓ 1 callersFunctionload_replica_traj
(filename: str)
eval_code/recons/relpose/evo_utils.py:107
↓ 1 callersFunctionload_seq_cameras
(example_path: str)
eval_code/recons/datasets/preprocess/prepare_re10k.py:7
↓ 1 callersFunctionload_sintel_traj
(gt_file)
eval_code/recons/relpose/evo_utils.py:136
↓ 1 callersFunctionload_split_info
Return the split json dict.
data/dataset_utils_vggt.py:61
↓ 1 callersFunctionmain
()
inference_recon.py:27
↓ 1 callersFunctionmain
(args)
data/preprocessing/process_tartanair.py:12
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/videodepth/eval.py:19
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/videodepth/infer.py:18
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/videodepth/infer_mp.py:20
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/monodepth/eval.py:20
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/monodepth/infer.py:22
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/monodepth/infer_mp.py:20
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/relpose/eval_angle.py:22
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/relpose/eval_dist.py:22
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/relpose/eval_dist_mp.py:21
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/relpose/eval_angle_mp.py:20
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/relpose/sampling.py:17
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/mv_recon/eval.py:23
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/mv_recon/sampling.py:17
↓ 1 callersFunctionmain
(hydra_cfg: DictConfig)
eval_code/recons/mv_recon/eval_mp.py:23
↓ 1 callersFunctionmain
()
train/joint_train_unified_model.py:357
↓ 1 callersFunctionmain
()
train/pretrain_unified_model.py:357
↓ 1 callersFunctionmake_device
Makes an actual torch.device object from the device specified as either a string or torch.device object. If the device is `cuda` without
modeling/g2vlm/device_utils.py:17
↓ 1 callersFunctionmap_np_dtype
(dtype)
eval_code/recons/models/moge/utils3d/numpy/rasterization.py:20
↓ 1 callersFunctionmask_aware_nearest_resize_numpy
Resize 2D map by nearest interpolation. Return the nearest neighbor index and mask of the resized map. ### Parameters - `inputs`: a sing
eval_code/recons/models/moge/utils/geometry_numpy.py:141
↓ 1 callersFunctionmasked_max
Similar to torch.max, but with mask
eval_code/recons/models/moge/utils3d/torch/utils.py:307
↓ 1 callersFunctionmasked_min
Similar to torch.min, but with mask
eval_code/recons/models/moge/utils3d/torch/utils.py:298
↓ 1 callersFunctionmatrix_to_quaternion
Convert 3x3 rotation matrix to quaternion (w, x, y, z) Args: rot_mat (torch.Tensor): shape (..., 3, 3), the rotation matrices to convert
eval_code/recons/models/moge/utils3d/torch/transforms.py:977
↓ 1 callersFunctionmax_pool_1d
(x: np.ndarray, kernel_size: int, stride: int, padding: int = 0, axis: int = -1)
eval_code/recons/models/moge/utils3d/numpy/utils.py:80
↓ 1 callersFunctionmax_pool_nd
(x: np.ndarray, kernel_size: Tuple[int,...], stride: Tuple[int,...], padding: Tuple[int,...], axis: Tuple[int,
eval_code/recons/models/moge/utils3d/numpy/utils.py:91
↓ 1 callersFunctionmodulate
Modulate the input tensor using scaling and shifting parameters.
eval_code/recons/models/vggt/heads/camera_head.py:157
↓ 1 callersFunctionnamed_apply
(fn: Callable, module: nn.Module, name="", depth_first=True, include_root=False)
modeling/pi3/models/dinov2/models/vision_transformer.py:27
↓ 1 callersFunctionnamed_apply
(fn: Callable, module: nn.Module, name="", depth_first=True, include_root=False)
eval_code/recons/models/moge/model/dinov2/models/vision_transformer.py:26
↓ 1 callersFunctionnamed_apply
(fn: Callable, module: nn.Module, name="", depth_first=True, include_root=False)
eval_code/recons/models/vggt/layers/vision_transformer.py:24
↓ 1 callersFunctionnamed_apply
(fn: Callable, module: nn.Module, name="", depth_first=True, include_root=False)
eval_code/recons/models/pi3/models/dinov2/models/vision_transformer.py:27
↓ 1 callersMethodnoraml_loss
(self, points, gt_points, mask)
modeling/pi3/models/pi3_loss.py:86
↓ 1 callersMethodnormalize_pred
(self, pred, gt)
modeling/pi3/models/pi3_loss.py:314
↓ 1 callersFunctionopencv_from_cameras_projection_RT
(R, T)
eval_code/recons/datasets/co3d_v2.py:56
↓ 1 callersFunctionopencv_to_colmap_intrinsics
Modify camera intrinsics to follow a different convention. Coordinates of the center of the top-left pixels are by default: - (0.5, 0.5)
modeling/pi3/utils/basic.py:55
↓ 1 callersMethodpack_sequence
(self, sample, sequence_status)
data/dataset_base_periter.py:386
↓ 1 callersMethodparse_row
(self, row)
data/interleave_datasets/interleave_dataset.py:162
↓ 1 callersFunctionpoints_to_normals
Calculate normal map from point map. Value range is [-1, 1]. Normal direction in OpenGL identity camera's coordinate system. Args: p
eval_code/recons/models/moge/utils3d/torch/utils.py:235
↓ 1 callersFunctionpoints_to_normals
Calculate normal map from point map. Value range is [-1, 1]. Normal direction in OpenGL identity camera's coordinate system. Args: p
eval_code/recons/models/moge/utils3d/numpy/utils.py:199
↓ 1 callersFunctionpoisson_equation
(width: int, height: int, wrap_x: bool = False, wrap_y: bool = False)
eval_code/recons/models/moge/utils/panorama.py:53
↓ 1 callersFunctionposition_grid_to_embed
Convert 2D position grid (HxWx2) to sinusoidal embeddings (HxWxC) Args: pos_grid: Tensor of shape (H, W, 2) containing 2D coordinate
eval_code/recons/models/vggt/heads/utils.py:11
↓ 1 callersMethodpredict_sky_mask
(self, imgs)
modeling/pi3/models/pi3_loss.py:61
↓ 1 callersMethodprepare_dino_images_none
(self, curr_kvlens, curr_rope, images, transforms, new_token_ids)
modeling/g2vlm/g2vlm.py:963
↓ 1 callersMethodprepare_gt
(self, gt)
modeling/pi3/models/pi3_loss.py:266
↓ 1 callersFunctionprepare_img
()
modeling/dinov2_with_registers/convert_dinov2_with_registers_to_hf.py:145
↓ 1 callersMethodprepare_segformer
(self)
modeling/pi3/models/pi3_loss.py:55
↓ 1 callersMethodprepare_start_tokens
(self, curr_kvlens, curr_rope, tokenizer, new_token_ids)
modeling/g2vlm/g2vlm.py:1135
↓ 1 callersMethodprepare_vit_images
(self, curr_kvlens, curr_rope, images, transforms, new_token_ids)
modeling/g2vlm/g2vlm.py:731
↓ 1 callersFunctionprocess_conversation
(images, conversation)
g2vlm_utils.py:119
↓ 1 callersFunctionprocess_poses
(co3d_dir, category, output_dir, min_quality)
eval_code/recons/datasets/preprocess/prepare_co3dv2.py:46
↓ 1 callersMethodprogram_texture
(self, n: int)
eval_code/recons/models/moge/utils3d/numpy/rasterization.py:74
↓ 1 callersMethodprovide
(self)
eval_code/recons/models/moge/utils/pipeline.py:188
↓ 1 callersMethodprune_heads
(self, heads: Set[int])
modeling/dinov2_with_registers/modeling_dinov2_with_registers.py:291
↓ 1 callersMethodprune_heads
(self, heads: Set[int])
modeling/g2vlm/dinov2_model.py:111
↓ 1 callersFunctionquat_to_mat
Quaternion Order: XYZW or say ijkr, scalar-last Convert rotations given as quaternions to rotation matrices. Args: quaternions:
eval_code/recons/models/vggt/utils/rotation.py:14
↓ 1 callersFunctionquaternion_invert
Given a quaternion representing rotation, get the quaternion representing its inverse. Args: quaternion: Quaternions as tensor o
modeling/g2vlm/rotation_utils.py:436
↓ 1 callersFunctionquaternion_to_axis_angle
Convert rotations given as quaternions to axis/angle. Args: quaternions: quaternions with real part first, as tensor of
modeling/g2vlm/rotation_utils.py:543
↓ 1 callersFunctionquaternion_to_axis_angle
Convert a batch of quaternions (w, x, y, z) to axis-angle representation (rotation vector) Args: quaternion (torch.Tensor): shape (..., 4
eval_code/recons/models/moge/utils3d/torch/transforms.py:946
↓ 1 callersFunctionquaternion_to_matrix
Converts a batch of quaternions (w, x, y, z) to rotation matrices Args: quaternion (np.ndarray): shape (..., 4), the quaternions to c
eval_code/recons/models/moge/utils3d/numpy/transforms.py:679
↓ 1 callersFunctionrandom_quaternions
Generate random quaternions representing rotations, i.e. versors with nonnegative real part. Args: n: Number of quaternions in a
modeling/g2vlm/rotation_utils.py:323
↓ 1 callersFunctionrandom_rotations
Generate random rotations as 3x3 rotation matrices. Args: n: Number of rotation matrices in a batch to return. dtype: Type t
modeling/g2vlm/rotation_utils.py:347
↓ 1 callersFunctionread_cameras_txt
(path)
eval_code/recons/datasets/preprocess/prepare_eth3d.py:12
↓ 1 callersFunctionread_images_txt
(path)
eval_code/recons/datasets/preprocess/prepare_eth3d.py:69
↓ 1 callersFunctionread_in_q_k_v
(state_dict, config)
modeling/dinov2_with_registers/convert_dinov2_with_registers_to_hf.py:126
↓ 1 callersMethodrecon
( self, tokenizer, new_token_ids, dino_image_transform, images, #this
modeling/g2vlm/g2vlm.py:1334
↓ 1 callersMethodrecon_for_eval
( self, tokenizer, new_token_ids, dino_image_transform, images, #this
modeling/g2vlm/g2vlm.py:1398
↓ 1 callersFunctionremove_unreferenced_vertices
Remove unreferenced vertices of a mesh. Unreferenced vertices are removed, and the face indices are updated accordingly. Args:
eval_code/recons/models/moge/utils3d/torch/mesh.py:407
↓ 1 callersFunctionrename_key
(dct, old, new)
modeling/dinov2_with_registers/convert_dinov2_with_registers_to_hf.py:120
↓ 1 callersMethodrot_ang_loss
Args: R: estimated rotation matrix [B, 3, 3] Rgt: ground-truth rotation matrix [B, 3, 3] Returns:
modeling/pi3/models/pi3_loss.py:199
↓ 1 callersMethodrot_pos_emb
(self, grid_thw)
modeling/qwen2vl/modeling_qwen2_vl_vit.py:396
↓ 1 callersMethodrot_pos_emb
(self, grid_thw)
modeling/qwen2vl/modeling_qwen2_vl.py:1019
← previousnext →601–700 of 2,118, ranked by callers