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

Adapt croco's DPTOutputAdapter implementation for dust3r: remove duplicated weigths, and fix forward for dust3r

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

Convenience class to aply the same operation to a whole set of images.

Convenience class to aply the same operation to a whole set of images.

Custom output class for ARCroco3DStereo.

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

Interpolation module.

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

DPT module for CroCo. by default, hooks_idx will be equal to: * for encoder-only: 4 equally spread layers * for encoder+decoder: last enco

Residual convolution module.

Track a series of values and provide access to smoothed values.

Batch sampler that groups indices from RandomSampler into batches.

Concatenation of several datasets

Random sampling under a constraint: each sample in the batch has the same feature, which is chosen randomly from a known pool of 'features' for ea

DPT output adapter. :param num_cahnnels: Number of output channels :param stride_level: tride level compared to the full-sized image.

scale and shift invariant loss

Feature fusion block.

Euclidean distance between 3d points

Euclidean distance between 3d points

Linear head for dust3r Each token outputs: - 16x16 3D points (+ confidence)

Linear head for dust3r Each token outputs: - 16x16 3D points (+ confidence)

Linear head for dust3r Each token outputs: - 16x16 3D points (+ confidence)

Artifically augmenting the size of a dataset.

just adding _init_weights + position getter compared to timm.models.layers.patch_embed.PatchEmbed

DPT module for dust3r, can return 3D points + confidence for all pixels

return positions of patches

return positions of patches

random masking

Artifically changing the size of a dataset.

Layer to compute the SSIM loss between a pair of images

scale invariant loss

Define all basic options. Usage: class MyDataset (BaseMultiViewDataset): def _get_views(self, idx, rng): # ov

Define all basic options. Usage: class MyDataset (BaseStereoViewDataset): def _get_views(self, idx, rng): # o

Dataset of outdoor street scenes, 5 images each time

Weighted regression by learned confidence. Assuming the input pixel_loss is a pixel-level regression loss. Principle: high-confid

Weighted regression by learned confidence. Assuming the input pixel_loss is a pixel-level regression loss. Principle: high-confid

a dataset that you can easily resize and combine. Examples: --------- 2 * dataset ==> duplicate each element 2x 10 @ dataset