---------- float ----------

Causal 3d convolusion.

---------- double ----------

---------- float ----------

Normlize image by given mean and std.

---------- double ----------

---------- float ----------

---------- float ----------

---------- double ----------

---------- float ----------

r""" A custom 3D causal convolution layer with feature caching support. This layer extends the standard Conv3D layer by ensuring causality in

Resize sample to given size (width, height).

Computes and stores the average and current value. Args: name (str): Name of the metric being tracked device (torch.device, option

Dataclass for depth estimation results. - relative_inv_depth: The estimated inverse depth map ([B,], H, W) in relative scale. One ha

MLP as used in Vision Transformer, MLP-Mixer and related networks

Replacement for mmcv.cnn.ConvModule to avoid mmcv dependency.

Prepare sample for usage as network input.

A video stream from a raw video stream, with processing applied.

Represents a matrix of sparse matrices. If the matrix is unspecified, it is assumed to be symmetric.

Causal self-attention with a single head.

Frame data from a single video frame. - raw_frame_idx: The index of the frame in the video. - rgb: The RGB image of the frame. The shape

r"""ConvNeXt A PyTorch impl of : `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf Args: in_chans (int):

Dataclass for depth estimation inputs. - rgb: The source image ([B,], H, W, 3), should be within 0-1 float. - video_frame_list: The list

Feature fusion block.

r""" A custom RMS normalization layer. Args: dim (int): The number of dimensions to normalize over. channel_first (bool, opti

Convert ndarrays in sample to Tensors.