Spaces:
Sleeping
Sleeping
| # Copyright (C) 2021-2024, Mindee. | |
| # This program is licensed under the Apache License 2.0. | |
| # See LICENSE or go to <https://opensource.org/licenses/Apache-2.0> for full license details. | |
| import math | |
| from typing import Tuple | |
| import torch | |
| from torch import nn | |
| __all__ = ["PatchEmbedding"] | |
| class PatchEmbedding(nn.Module): | |
| """Compute 2D patch embeddings with cls token and positional encoding""" | |
| def __init__(self, input_shape: Tuple[int, int, int], embed_dim: int, patch_size: Tuple[int, int]) -> None: | |
| super().__init__() | |
| channels, height, width = input_shape | |
| self.patch_size = patch_size | |
| self.interpolate = True if patch_size[0] == patch_size[1] else False | |
| self.grid_size = tuple([s // p for s, p in zip((height, width), self.patch_size)]) | |
| self.num_patches = self.grid_size[0] * self.grid_size[1] | |
| self.cls_token = nn.Parameter(torch.randn(1, 1, embed_dim)) | |
| self.positions = nn.Parameter(torch.randn(1, self.num_patches + 1, embed_dim)) | |
| self.projection = nn.Conv2d(channels, embed_dim, kernel_size=self.patch_size, stride=self.patch_size) | |
| def interpolate_pos_encoding(self, embeddings: torch.Tensor, height: int, width: int) -> torch.Tensor: | |
| """100 % borrowed from: | |
| https://github.com/huggingface/transformers/blob/main/src/transformers/models/vit/modeling_vit.py | |
| This method allows to interpolate the pre-trained position encodings, to be able to use the model on higher | |
| resolution images. | |
| Source: | |
| https://github.com/facebookresearch/dino/blob/de9ee3df6cf39fac952ab558447af1fa1365362a/vision_transformer.py | |
| """ | |
| num_patches = embeddings.shape[1] - 1 | |
| num_positions = self.positions.shape[1] - 1 | |
| if num_patches == num_positions and height == width: | |
| return self.positions | |
| class_pos_embed = self.positions[:, 0] | |
| patch_pos_embed = self.positions[:, 1:] | |
| dim = embeddings.shape[-1] | |
| h0 = float(height // self.patch_size[0]) | |
| w0 = float(width // self.patch_size[1]) | |
| # we add a small number to avoid floating point error in the interpolation | |
| # see discussion at https://github.com/facebookresearch/dino/issues/8 | |
| h0, w0 = h0 + 0.1, w0 + 0.1 | |
| patch_pos_embed = patch_pos_embed.reshape(1, int(math.sqrt(num_positions)), int(math.sqrt(num_positions)), dim) | |
| patch_pos_embed = patch_pos_embed.permute(0, 3, 1, 2) | |
| patch_pos_embed = nn.functional.interpolate( | |
| patch_pos_embed, | |
| scale_factor=(h0 / math.sqrt(num_positions), w0 / math.sqrt(num_positions)), | |
| mode="bilinear", | |
| align_corners=False, | |
| recompute_scale_factor=True, | |
| ) | |
| assert int(h0) == patch_pos_embed.shape[-2], "height of interpolated patch embedding doesn't match" | |
| assert int(w0) == patch_pos_embed.shape[-1], "width of interpolated patch embedding doesn't match" | |
| patch_pos_embed = patch_pos_embed.permute(0, 2, 3, 1).view(1, -1, dim) | |
| return torch.cat((class_pos_embed.unsqueeze(0), patch_pos_embed), dim=1) | |
| def forward(self, x: torch.Tensor) -> torch.Tensor: | |
| B, C, H, W = x.shape | |
| assert H % self.patch_size[0] == 0, "Image height must be divisible by patch height" | |
| assert W % self.patch_size[1] == 0, "Image width must be divisible by patch width" | |
| # patchify image | |
| patches = self.projection(x).flatten(2).transpose(1, 2) | |
| cls_tokens = self.cls_token.expand(B, -1, -1) # (batch_size, 1, d_model) | |
| # concate cls_tokens to patches | |
| embeddings = torch.cat([cls_tokens, patches], dim=1) # (batch_size, num_patches + 1, d_model) | |
| # add positions to embeddings | |
| if self.interpolate: | |
| embeddings += self.interpolate_pos_encoding(embeddings, H, W) | |
| else: | |
| embeddings += self.positions | |
| return embeddings # (batch_size, num_patches + 1, d_model) | |