models_rc.py

# --------------------------------------------------------------------------------
# Exploring the Role of Mean Teachers in Self-supervised Masked Auto-Encoders (ICLR'23)
# Copyright (c) 2022 Electronics and Telecommunications Research Institute (ETRI).
# All Rights Reserved.
# Written by Youngwan Lee
# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.
# --------------------------------------------------------------------------------
# Modified from MAE (https://github.com/facebookresearch/mae)
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
# --------------------------------------------------------------------------------
# References:
# MAE: https://github.com/facebookresearch/mae
# DeiT: https://github.com/facebookresearch/deit
# BEiT: https://github.com/microsoft/unilm/tree/master/beit
# --------------------------------------------------------------------------------

"""Reconstruction-Consistent Masked Auto-Encoder"""


from functools import partial

import torch
import torch.nn as nn

from timm.models.vision_transformer import PatchEmbed, Block

from util.pos_embed import get_2d_sincos_pos_embed


class MaskedAutoencoderViT(nn.Module):
    """Masked Autoencoder with VisionTransformer backbone"""

    def __init__(
        self,
        img_size=224,
        patch_size=16,
        in_chans=3,
        embed_dim=1024,
        depth=24,
        num_heads=16,
        decoder_embed_dim=512,
        decoder_depth=8,
        decoder_num_heads=16,
        mlp_ratio=4.0,
        norm_layer=nn.LayerNorm,
        norm_pix_loss=False,
        compute_loss=True,
    ):
        super().__init__()

        # --------------------------------------------------------------------------
        # MAE encoder specifics
        self.patch_embed = PatchEmbed(
            img_size, patch_size, in_chans, embed_dim)
        num_patches = self.patch_embed.num_patches

        self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
        self.pos_embed = nn.Parameter(
            torch.zeros(1, num_patches + 1, embed_dim), requires_grad=False
        )  # fixed sin-cos embedding

        self.blocks = nn.ModuleList(
            [
                Block(
                    embed_dim,
                    num_heads,
                    mlp_ratio,
                    qkv_bias=True,
                    qk_scale=None,
                    norm_layer=norm_layer,
                )
                for i in range(depth)
            ]
        )
        self.norm = norm_layer(embed_dim)
        # --------------------------------------------------------------------------

        # --------------------------------------------------------------------------
        # MAE decoder specifics
        self.decoder_embed = nn.Linear(embed_dim, decoder_embed_dim, bias=True)

        self.mask_token = nn.Parameter(torch.zeros(1, 1, decoder_embed_dim))

        self.decoder_pos_embed = nn.Parameter(
            torch.zeros(
                1,
                num_patches + 1,
                decoder_embed_dim),
            requires_grad=False)  # fixed sin-cos embedding

        self.decoder_blocks = nn.ModuleList(
            [
                Block(
                    decoder_embed_dim,
                    decoder_num_heads,
                    mlp_ratio,
                    qkv_bias=True,
                    qk_scale=None,
                    norm_layer=norm_layer,
                )
                for i in range(decoder_depth)
            ]
        )

        self.decoder_norm = norm_layer(decoder_embed_dim)
        self.decoder_pred = nn.Linear(
            decoder_embed_dim, patch_size ** 2 * in_chans, bias=True
        )  # decoder to patch
        # --------------------------------------------------------------------------

        self.norm_pix_loss = norm_pix_loss

        self.compute_loss = compute_loss

        self.initialize_weights()

    def initialize_weights(self):
        """initialize (and freeze) pos_embed by sin-cos embedding"""

        pos_embed = get_2d_sincos_pos_embed(
            self.pos_embed.shape[-1],
            int(self.patch_embed.num_patches ** 0.5),
            cls_token=True,
        )
        self.pos_embed.data.copy_(
            torch.from_numpy(pos_embed).float().unsqueeze(0))

        decoder_pos_embed = get_2d_sincos_pos_embed(
            self.decoder_pos_embed.shape[-1],
            int(self.patch_embed.num_patches ** 0.5),
            cls_token=True,
        )
        self.decoder_pos_embed.data.copy_(
            torch.from_numpy(decoder_pos_embed).float().unsqueeze(0)
        )

        # initialize patch_embed like nn.Linear (instead of nn.Conv2d)
        w = self.patch_embed.proj.weight.data
        torch.nn.init.xavier_uniform_(w.view([w.shape[0], -1]))

        # timm's trunc_normal_(std=.02) is effectively normal_(std=0.02) as
        # cutoff is too big (2.)
        torch.nn.init.normal_(self.cls_token, std=0.02)
        torch.nn.init.normal_(self.mask_token, std=0.02)

        # initialize nn.Linear and nn.LayerNorm
        self.apply(self._init_weights)

    def _init_weights(self, m):
        if isinstance(m, nn.Linear):
            # we use xavier_uniform following official JAX ViT:
            torch.nn.init.xavier_uniform_(m.weight)
            if isinstance(m, nn.Linear) and m.bias is not None:
                nn.init.constant_(m.bias, 0)
        elif isinstance(m, nn.LayerNorm):
            nn.init.constant_(m.bias, 0)
            nn.init.constant_(m.weight, 1.0)

    def patchify(self, imgs):
        """
        imgs: (N, 3, H, W)
        x: (N, L, patch_size**2 *3)
        """
        p = self.patch_embed.patch_size[0]
        assert imgs.shape[2] == imgs.shape[3] and imgs.shape[2] % p == 0

        h = w = imgs.shape[2] // p
        x = imgs.reshape(shape=(imgs.shape[0], 3, h, p, w, p))
        x = torch.einsum("nchpwq->nhwpqc", x)
        x = x.reshape(shape=(imgs.shape[0], h * w, p ** 2 * 3))
        return x

    def unpatchify(self, x):
        """
        x: (N, L, patch_size**2 *3)
        imgs: (N, 3, H, W)
        """
        p = self.patch_embed.patch_size[0]
        h = w = int(x.shape[1] ** 0.5)
        assert h * w == x.shape[1]

        x = x.reshape(shape=(x.shape[0], h, w, p, p, 3))
        x = torch.einsum("nhwpqc->nchpwq", x)
        imgs = x.reshape(shape=(x.shape[0], 3, h * p, h * p))
        return imgs

    def random_masking(self, x, mask_ratio, ids_shuffle=None):
        """
        Perform per-sample random masking by per-sample shuffling.
        Per-sample shuffling is done by argsort random noise.
        x: [N, L, D], sequence
        """
        N, L, D = x.shape  # batch, length, dim
        len_keep = int(L * (1 - mask_ratio))

        if ids_shuffle is None:
            noise = torch.rand(N, L, device=x.device)  # noise in [0, 1]
            # sort noise for each sample
            # ascend: small is keep, large is remove
            ids_shuffle = torch.argsort(noise, dim=1)
        ids_restore = torch.argsort(ids_shuffle, dim=1)

        # keep the first subset
        ids_keep = ids_shuffle[:, :len_keep]
        x_masked = torch.gather(
            x, dim=1, index=ids_keep.unsqueeze(-1).repeat(1, 1, D))

        # generate the binary mask: 0 is keep, 1 is remove
        mask = torch.ones([N, L], device=x.device)
        mask[:, :len_keep] = 0
        # unshuffle to get the binary mask
        mask = torch.gather(mask, dim=1, index=ids_restore)

        return x_masked, mask, ids_restore, ids_shuffle

    def forward_encoder(self, x, mask_ratio, ids_shuffle=None):
        """
        forward step in encoder
        """
        # embed patches
        # x : (B, 14*14, embed_dim)
        x = self.patch_embed(x)

        # add pos embed w/o cls token
        x = x + self.pos_embed[:, 1:, :]

        # masking: length -> length * mask_ratio
        # ywlee
        # x : (B, 49 = length * mask_ratio, embed_dim)
        #    e.g., (B, 49, 384)
        # mask : (B, length=14*14)
        # ids_restore : (B, length=14*14)
        x, mask, ids_restore, ids_shuffle = self.random_masking(
            x, mask_ratio, ids_shuffle
        )

        # append cls token
        cls_token = self.cls_token + self.pos_embed[:, :1, :]
        cls_tokens = cls_token.expand(x.shape[0], -1, -1)
        x = torch.cat((cls_tokens, x), dim=1)

        # apply Transformer blocks
        for blk in self.blocks:
            x = blk(x)
        x = self.norm(x)

        return x, mask, ids_restore, ids_shuffle

    def forward_decoder(self, x, ids_restore):
        """
        forward step in decoder
        """
        # embed tokens
        # encoder's embed_dim -> decoder's embed_dim
        # (B, length, embed_dim) -> (B, length=14*14, decoder_embed_dim)
        # (2, 49+1, 384) -> (2, 49+1, 512)
        x = self.decoder_embed(x)

        # append mask tokens to sequence
        # mask_tokens : (B, length*mask_ratio, decoder_embed_dim)
        # x_ : (B, length, decoder_embed_dim)
        # x  : (B, length+1, decoder_embed_dim)
        mask_tokens = self.mask_token.repeat(
            x.shape[0], ids_restore.shape[1] + 1 - x.shape[1], 1
        )
        x_ = torch.cat([x[:, 1:, :], mask_tokens], dim=1)  # no cls token
        x_ = torch.gather(
            x_, dim=1, index=ids_restore.unsqueeze(-1).repeat(1, 1, x.shape[2])
        )  # unshuffle
        x = torch.cat([x[:, :1, :], x_], dim=1)  # append cls token

        # add pos embed
        x = x + self.decoder_pos_embed

        # apply Transformer blocks
        for blk in self.decoder_blocks:
            x = blk(x)
        x = self.decoder_norm(x)

        # predictor projection
        # x : (B, length+1, decoder_embed_dim) -> (B, length+1, patch_size**2 *
        # in_chans)
        x = self.decoder_pred(x)

        # remove cls token
        x = x[:, 1:, :]

        return x

    def forward_loss(self, imgs, pred, mask):
        """
        imgs: [N, 3, H, W]
        pred: [N, L, p*p*3]
        mask: [N, L], 0 is keep, 1 is remove,
        """
        # target : [N, L, p*p*3]
        # e.g., [2, 196, 768]
        target = self.patchify(imgs)
        if self.norm_pix_loss:
            mean = target.mean(dim=-1, keepdim=True)
            var = target.var(dim=-1, keepdim=True)
            target = (target - mean) / (var + 1.0e-6) ** 0.5

        loss = (pred - target) ** 2
        loss = loss.mean(dim=-1)  # [N, L], mean loss per patch

        loss = (loss * mask).sum() / mask.sum()  # mean loss on removed patches
        return loss

    def forward(self, imgs, mask_ratio=0.75, ids_shuffle=None):
        """
        overall forward step
        """
        latent, mask, ids_restore, ids_shuffle = self.forward_encoder(
            imgs, mask_ratio, ids_shuffle
        )
        pred = self.forward_decoder(latent, ids_restore)  # [N, L, p*p*3]
        if self.compute_loss:
            loss = self.forward_loss(imgs, pred, mask)
            return loss, pred, mask, ids_shuffle
        else:  # for teacher network
            return pred


def rc_vit_tiny_patch16_dec128d2b(**kwargs):
    """ rc-mae with vit-tiny """
    model = MaskedAutoencoderViT(
        patch_size=16,
        embed_dim=192,
        depth=12,
        num_heads=3,
        decoder_embed_dim=128,
        decoder_depth=2,
        decoder_num_heads=4,
        mlp_ratio=4,
        norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs
    )
    return model


def rc_vit_small_patch16_dec512d8b(**kwargs):
    """ rc-mae with vit-small """
    model = MaskedAutoencoderViT(
        patch_size=16,
        embed_dim=384,
        depth=12,
        num_heads=6,
        decoder_embed_dim=512,
        decoder_depth=8,
        decoder_num_heads=16,
        mlp_ratio=4,
        norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs
    )
    return model


def rc_vit_small_patch16_dec256d4b(**kwargs):
    """ rc-mae with vit-small with 256-dim & 4 blocks """
    model = MaskedAutoencoderViT(
        patch_size=16,
        embed_dim=384,
        depth=12,
        num_heads=6,
        decoder_embed_dim=256,
        decoder_depth=4,
        decoder_num_heads=8,
        mlp_ratio=4,
        norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs
    )
    return model


def rc_vit_base_patch16_dec512d8b(**kwargs):
    """ rc-mae with vit-base """
    model = MaskedAutoencoderViT(
        patch_size=16,
        embed_dim=768,
        depth=12,
        num_heads=12,
        decoder_embed_dim=512,
        decoder_depth=8,
        decoder_num_heads=16,
        mlp_ratio=4,
        norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs
    )
    return model


def rc_vit_large_patch16_dec512d8b(**kwargs):
    """ rc-mae with vit-large """

    model = MaskedAutoencoderViT(
        patch_size=16,
        embed_dim=1024,
        depth=24,
        num_heads=16,
        decoder_embed_dim=512,
        decoder_depth=8,
        decoder_num_heads=16,
        mlp_ratio=4,
        norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs
    )
    return model


def rc_vit_huge_patch14_dec512d8b(**kwargs):
    """ rc-mae with vit-huge """

    model = MaskedAutoencoderViT(
        patch_size=14,
        embed_dim=1280,
        depth=32,
        num_heads=16,
        decoder_embed_dim=512,
        decoder_depth=8,
        decoder_num_heads=16,
        mlp_ratio=4,
        norm_layer=partial(nn.LayerNorm, eps=1e-6),
        **kwargs
    )
    return model


# set recommended archs
# ywlee add
# decoder: 512 dim, 8 blocksmae_vit_small_patch16_dec512d8b
rc_vit_tiny_patch16 = rc_vit_tiny_patch16_dec128d2b
# mae_vit_small_patch16 = mae_vit_small_patch16_dec512d8b  # decoder: 512
# dim, 8 blocksmae_vit_small_patch16_dec512d8b
# decoder: 512 dim, 8 blocksmae_vit_small_patch16_dec512d8b
rc_vit_small_patch16 = rc_vit_small_patch16_dec256d4b
# ywlee end
rc_vit_base_patch16 = rc_vit_base_patch16_dec512d8b  # decoder: 512 dim, 8 blocks
rc_vit_large_patch16 = rc_vit_large_patch16_dec512d8b  # decoder: 512 dim, 8 blocks
rc_vit_huge_patch14 = rc_vit_huge_patch14_dec512d8b  # decoder: 512 dim, 8 blocks


if __name__ == "__main__":
    """
    model spec check
    """
    student = rc_vit_small_patch16_dec512d8b(compute_loss=True)
    teacher = rc_vit_small_patch16_dec512d8b(compute_loss=False)
    student.eval()
    teacher.eval()
    inputs = torch.randn(2, 3, 224, 224)
    s_loss, s_pred, s_mask, ids_shuffle = student(inputs, mask_ratio=0.75)
    t_pred, t_mask = teacher(inputs, mask_ratio=0.75, ids_shuffle=ids_shuffle)