add backbone vitpose

myeasymocap/backbone/basetopdown.py

@@ -49,6 +49,37 @@ def gen_trans_from_patch_cv(c_x, c_y, src_width, src_height, dst_width, dst_heig
 
     return trans, inv_trans
 
+# TODO: add UDP
+def get_warp_matrix(theta, size_input, size_dst, size_target):
+    """Calculate the transformation matrix under the constraint of unbiased.
+    Paper ref: Huang et al. The Devil is in the Details: Delving into Unbiased
+    Data Processing for Human Pose Estimation (CVPR 2020).
+
+    Args:
+        theta (float): Rotation angle in degrees.
+        size_input (np.ndarray): Size of input image [w, h].
+        size_dst (np.ndarray): Size of output image [w, h].
+        size_target (np.ndarray): Size of ROI in input plane [w, h].
+
+    Returns:
+        np.ndarray: A matrix for transformation.
+    """
+    theta = np.deg2rad(theta)
+    matrix = np.zeros((2, 3), dtype=np.float32)
+    scale_x = size_dst[0] / size_target[0]
+    scale_y = size_dst[1] / size_target[1]
+    matrix[0, 0] = math.cos(theta) * scale_x
+    matrix[0, 1] = -math.sin(theta) * scale_x
+    matrix[0, 2] = scale_x * (-0.5 * size_input[0] * math.cos(theta) +
+                              0.5 * size_input[1] * math.sin(theta) +
+                              0.5 * size_target[0])
+    matrix[1, 0] = math.sin(theta) * scale_y
+    matrix[1, 1] = math.cos(theta) * scale_y
+    matrix[1, 2] = scale_y * (-0.5 * size_input[0] * math.sin(theta) -
+                              0.5 * size_input[1] * math.cos(theta) +
+                              0.5 * size_target[1])
+    return matrix
+
 def generate_patch_image_cv(cvimg, c_x, c_y, bb_width, bb_height, patch_width, patch_height, do_flip, scale, rot):
 
     trans, inv_trans = gen_trans_from_patch_cv(c_x, c_y, bb_width, bb_height, patch_width, patch_height, scale, rot, inv=False)
@@ -75,8 +106,8 @@ def get_single_image_crop_demo(image, bbox, scale=1.2, crop_size=224,
     )
     if fliplr:
         crop_image = cv2.flip(crop_image, 1)
-    # cv2.imwrite('debug_crop.jpg', crop_image)
-    # import ipdb; ipdb.set_trace()
+    # cv2.imwrite('debug_crop.jpg', crop_image[:,:,::-1])
+    # cv2.imwrite('debug_crop_full.jpg', image[:,:,::-1])
     crop_image = crop_image.transpose(2,0,1)
     mean1=np.array(mean, dtype=np.float32).reshape(3,1,1)
     std1= np.array(std, dtype=np.float32).reshape(3,1,1)
@@ -123,6 +154,14 @@ class BaseTopDownModel(nn.Module):
             squeeze = True
         # TODO: 兼容多张图片的
         bbox = xyxy2ccwh(bbox)
+        # convert the bbox to the aspect of input bbox
+        aspect_ratio = self.crop_size[1] / self.crop_size[0]
+        w, h = bbox[:, 2], bbox[:, 3]
+        # 如果height大于w*ratio，那么增大w
+        flag = h > aspect_ratio * w
+        bbox[flag, 2] = h[flag] / aspect_ratio
+        # 否则增大h
+        bbox[~flag, 3] = w[~flag] * aspect_ratio
         inputs = []
         inv_trans_ = []
         for i in range(bbox.shape[0]):
@@ -141,6 +180,15 @@ class BaseTopDownModel(nn.Module):
             )
             inputs.append(norm_img)
             inv_trans_.append(inv_trans)
+        if False:
+            vis = np.hstack(inputs)
+            mean, std = np.array(self.mean), np.array(self.std)
+            mean = mean.reshape(3, 1, 1)
+            std = std.reshape(3, 1, 1)
+            vis = (vis * std) + mean
+            vis = vis.transpose(1, 2, 0)
+            vis = (vis[:, :, ::-1] * 255).astype(np.uint8)
+            cv2.imwrite('debug_crop.jpg', vis)
         inputs = np.stack(inputs)
         inv_trans_ = np.stack(inv_trans_)
         inputs = torch.FloatTensor(inputs).to(self.device)
@@ -168,17 +216,30 @@ class BaseTopDownModelCache(BaseTopDownModel):
         super().__init__(**kwargs)
         self.name = name
     
-    def __call__(self, bbox, images, imgname, flips=None):
+    def cachename(self, imgname):
         basename = os.sep.join(imgname.split(os.sep)[-2:])
         cachename = join(self.output, self.name, basename.replace('.jpg', '.pkl'))
+        return cachename
+
+    def dump(self, cachename, output):
         os.makedirs(os.path.dirname(cachename), exist_ok=True)
+        with open(cachename, 'wb') as f:
+            pickle.dump(output, f)
+        return output
+    
+    def load(self, cachename):
+        with open(cachename, 'rb') as f:
+            output = pickle.load(f)
+        return output
+
+    def __call__(self, bbox, images, imgname, flips=None):
+        cachename = self.cachename(imgname)
         if os.path.exists(cachename):
-            with open(cachename, 'rb') as f:
-                output = pickle.load(f)
+            output = self.load(cachename)
         else:
             output = self.infer(images, bbox, to_numpy=True, flips=flips)
-            with open(cachename, 'wb') as f:
-                pickle.dump(output, f)
+            output = self.dump(cachename, output)
+
         ret = {
             'params': output
         }

myeasymocap/backbone/hrnet/myhrnet.py

@@ -51,9 +51,12 @@ def coco17tobody25(points2d):
     return kpts
 
 class MyHRNet(BaseTopDownModelCache):
-    def __init__(self, ckpt):
-        super().__init__(name='hand2d', bbox_scale=1.25, res_input=[288, 384])
-        model = HRNet(48, 17, 0.1)
+    def __init__(self, ckpt, single_person=True, num_joints=17, name='keypoints2d'):
+        super().__init__(name, bbox_scale=1.25, res_input=[288, 384])
+        # 如果启用，那么将每个视角最多保留一个，并且squeeze and stack
+        self.single_person = single_person
+        model = HRNet(48, num_joints, 0.1)
+        self.num_joints = num_joints
         if not os.path.exists(ckpt) and ckpt.endswith('pose_hrnet_w48_384x288.pth'):
             url = "11ezQ6a_MxIRtj26WqhH3V3-xPI3XqYAw"
             text = '''Download `models/pytorch/pose_coco/pose_hrnet_w48_384x288.pth` from (OneDrive)[https://1drv.ms/f/s!AhIXJn_J-blW231MH2krnmLq5kkQ],
@@ -109,20 +112,23 @@ class MyHRNet(BaseTopDownModelCache):
         for nv in range(nViews):
             _bbox = bbox[nv]
             if _bbox.shape[0] == 0:
-                kpts_all.append(np.zeros((17, 3)))
-                continue
-            img = images[nv]
-            # TODO: add flip test
-            out = super().__call__(_bbox, img, imgnames[nv])
-            output = out['params']['output']
-            kpts = self.get_max_preds(output)
-            kpts_ori = self.batch_affine_transform(kpts, out['params']['inv_trans'])
-            kpts = np.concatenate([kpts_ori, kpts[..., -1:]], axis=-1)
+                if self.single_person:
+                    kpts = np.zeros((1, self.num_joints, 3))
+                else:
+                    kpts = np.zeros((_bbox.shape[0], self.num_joints, 3))
+            else:
+                img = images[nv]
+                # TODO: add flip test
+                out = super().__call__(_bbox, img, imgnames[nv])
+                output = out['params']['output']
+                kpts = self.get_max_preds(output)
+                kpts_ori = self.batch_affine_transform(kpts, out['params']['inv_trans'])
+                kpts = np.concatenate([kpts_ori, kpts[..., -1:]], axis=-1)
             kpts = coco17tobody25(kpts)
-            if len(kpts.shape) == 3:
-                kpts = kpts[0]
             kpts_all.append(kpts)
-        kpts_all = np.stack(kpts_all)
+        if self.single_person:
+            kpts_all = [k[0] for k in kpts_all]
+            kpts_all = np.stack(kpts_all)
         if squeeze:
             kpts_all = kpts_all[0]
         return {

myeasymocap/backbone/topdown_keypoints.py

@@ -0,0 +1,94 @@
+import math
+import numpy as np
+
+def get_max_preds(batch_heatmaps):
+    '''
+    get predictions from score maps
+    heatmaps: numpy.ndarray([batch_size, num_joints, height, width])
+    '''
+    assert isinstance(batch_heatmaps, np.ndarray), \
+        'batch_heatmaps should be numpy.ndarray'
+    assert batch_heatmaps.ndim == 4, 'batch_images should be 4-ndim: {}'.format(batch_heatmaps.shape)
+
+    batch_size = batch_heatmaps.shape[0]
+    num_joints = batch_heatmaps.shape[1]
+    width = batch_heatmaps.shape[3]
+    heatmaps_reshaped = batch_heatmaps.reshape((batch_size, num_joints, -1))
+    idx = np.argmax(heatmaps_reshaped, 2)
+    maxvals = np.amax(heatmaps_reshaped, 2)
+
+    maxvals = maxvals.reshape((batch_size, num_joints, 1))
+    idx = idx.reshape((batch_size, num_joints, 1))
+
+    preds = np.tile(idx, (1, 1, 2)).astype(np.float32)
+
+    preds[:, :, 0] = (preds[:, :, 0]) % width
+    preds[:, :, 1] = np.floor((preds[:, :, 1]) / width)
+
+    pred_mask = np.tile(np.greater(maxvals, 0.0), (1, 1, 2))
+    pred_mask = pred_mask.astype(np.float32)
+
+    preds *= pred_mask
+    return preds, maxvals
+
+COCO17_IN_BODY25 = [0,16,15,18,17,5,2,6,3,7,4,12,9,13,10,14,11]
+pairs = [[1, 8], [1, 2], [1, 5], [2, 3], [3, 4], [5, 6], [6, 7], [8, 9], [9, 10], [10, 11], [8, 12], [12, 13], [13, 14], [1, 0], [0,15], [15,17], [0,16], [16,18], [14,19], [19,20], [14,21], [11,22], [22,23], [11,24]]
+def coco17tobody25(points2d):
+    kpts = np.zeros((points2d.shape[0], 25, 3))
+    kpts[:, COCO17_IN_BODY25, :2] = points2d[:, :, :2]
+    kpts[:, COCO17_IN_BODY25, 2:3] = points2d[:, :, 2:3]
+    kpts[:, 8, :2] = kpts[:, [9, 12], :2].mean(axis=1)
+    kpts[:, 8, 2] = kpts[:, [9, 12], 2].min(axis=1)
+    kpts[:, 1, :2] = kpts[:, [2, 5], :2].mean(axis=1)
+    kpts[:, 1, 2] = kpts[:, [2, 5], 2].min(axis=1)
+    # 需要交换一下
+    # kpts = kpts[:, :, [1,0,2]]
+    return kpts
+
+def coco23tobody25(points2d):
+    kpts = coco17tobody25(points2d[:, :17])
+    kpts[:, [19, 20, 21, 22, 23, 24]] = points2d[:, [17, 18, 19, 20, 21, 22]]
+    return kpts
+
+class BaseKeypoints():
+    @staticmethod
+    def get_max_preds(batch_heatmaps):
+        coords, maxvals = get_max_preds(batch_heatmaps)
+
+        heatmap_height = batch_heatmaps.shape[2]
+        heatmap_width = batch_heatmaps.shape[3]
+
+        # post-processing
+        if True:
+            for n in range(coords.shape[0]):
+                for p in range(coords.shape[1]):
+                    hm = batch_heatmaps[n][p]
+                    px = int(math.floor(coords[n][p][0] + 0.5))
+                    py = int(math.floor(coords[n][p][1] + 0.5))
+                    if 1 < px < heatmap_width-1 and 1 < py < heatmap_height-1:
+                        diff = np.array(
+                            [
+                                hm[py][px+1] - hm[py][px-1],
+                                hm[py+1][px]-hm[py-1][px]
+                            ]
+                        )
+                        coords[n][p] += np.sign(diff) * .25
+        coords = coords.astype(np.float32) * 4
+        pred = np.dstack((coords, maxvals))
+        return pred
+    
+    @staticmethod
+    def batch_affine_transform(points, trans):
+        # points: (Bn, J, 2), trans: (Bn, 2, 3)
+        points = np.dstack((points[..., :2], np.ones((*points.shape[:-1], 1))))
+        out = np.matmul(points, trans.swapaxes(-1, -2))
+        return out
+    
+    @staticmethod
+    def coco17tobody25(points2d):
+        return coco17tobody25(points2d)
+    
+    @staticmethod
+    def coco23tobody25(points2d):
+        return coco23tobody25(points2d)
+    

myeasymocap/backbone/vitpose/layers.py

@@ -0,0 +1,98 @@
+import torch
+import math
+import collections.abc
+from itertools import repeat
+import warnings
+import torch.nn as nn
+
+def drop_path(x, drop_prob: float = 0., training: bool = False, scale_by_keep: bool = True):
+    """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
+
+    This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
+    the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
+    See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for
+    changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use
+    'survival rate' as the argument.
+
+    """
+    if drop_prob == 0. or not training:
+        return x
+    keep_prob = 1 - drop_prob
+    shape = (x.shape[0],) + (1,) * (x.ndim - 1)  # work with diff dim tensors, not just 2D ConvNets
+    random_tensor = x.new_empty(shape).bernoulli_(keep_prob)
+    if keep_prob > 0.0 and scale_by_keep:
+        random_tensor.div_(keep_prob)
+    return x * random_tensor
+
+# From PyTorch internals
+def _ntuple(n):
+    def parse(x):
+        if isinstance(x, collections.abc.Iterable) and not isinstance(x, str):
+            return tuple(x)
+        return tuple(repeat(x, n))
+    return parse
+
+
+to_1tuple = _ntuple(1)
+to_2tuple = _ntuple(2)
+
+def _trunc_normal_(tensor, mean, std, a, b):
+    # Cut & paste from PyTorch official master until it's in a few official releases - RW
+    # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf
+    def norm_cdf(x):
+        # Computes standard normal cumulative distribution function
+        return (1. + math.erf(x / math.sqrt(2.))) / 2.
+
+    if (mean < a - 2 * std) or (mean > b + 2 * std):
+        warnings.warn("mean is more than 2 std from [a, b] in nn.init.trunc_normal_. "
+                      "The distribution of values may be incorrect.",
+                      stacklevel=2)
+
+    # Values are generated by using a truncated uniform distribution and
+    # then using the inverse CDF for the normal distribution.
+    # Get upper and lower cdf values
+    l = norm_cdf((a - mean) / std)
+    u = norm_cdf((b - mean) / std)
+
+    # Uniformly fill tensor with values from [l, u], then translate to
+    # [2l-1, 2u-1].
+    tensor.uniform_(2 * l - 1, 2 * u - 1)
+
+    # Use inverse cdf transform for normal distribution to get truncated
+    # standard normal
+    tensor.erfinv_()
+
+    # Transform to proper mean, std
+    tensor.mul_(std * math.sqrt(2.))
+    tensor.add_(mean)
+
+    # Clamp to ensure it's in the proper range
+    tensor.clamp_(min=a, max=b)
+    return tensor
+
+
+def trunc_normal_(tensor, mean=0., std=1., a=-2., b=2.):
+    # type: (Tensor, float, float, float, float) -> Tensor
+    r"""Fills the input Tensor with values drawn from a truncated
+    normal distribution. The values are effectively drawn from the
+    normal distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)`
+    with values outside :math:`[a, b]` redrawn until they are within
+    the bounds. The method used for generating the random values works
+    best when :math:`a \leq \text{mean} \leq b`.
+
+    NOTE: this impl is similar to the PyTorch trunc_normal_, the bounds [a, b] are
+    applied while sampling the normal with mean/std applied, therefore a, b args
+    should be adjusted to match the range of mean, std args.
+
+    Args:
+        tensor: an n-dimensional `torch.Tensor`
+        mean: the mean of the normal distribution
+        std: the standard deviation of the normal distribution
+        a: the minimum cutoff value
+        b: the maximum cutoff value
+    Examples:
+        >>> w = torch.empty(3, 5)
+        >>> nn.init.trunc_normal_(w)
+    """
+    with torch.no_grad():
+        return _trunc_normal_(tensor, mean, std, a, b)

myeasymocap/backbone/vitpose/vit_moe.py

@@ -0,0 +1,607 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import os
+import numpy as np
+import torch
+from functools import partial
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as checkpoint
+
+from .layers import drop_path, to_2tuple, trunc_normal_
+
+def drop_path(x, drop_prob: float = 0., training: bool = False, scale_by_keep: bool = True):
+    """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
+
+    This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
+    the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
+    See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for
+    changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use
+    'survival rate' as the argument.
+
+    """
+    if drop_prob == 0. or not training:
+        return x
+    keep_prob = 1 - drop_prob
+    shape = (x.shape[0],) + (1,) * (x.ndim - 1)  # work with diff dim tensors, not just 2D ConvNets
+    random_tensor = x.new_empty(shape).bernoulli_(keep_prob)
+    if keep_prob > 0.0 and scale_by_keep:
+        random_tensor.div_(keep_prob)
+    return x * random_tensor
+
+def get_abs_pos(abs_pos, h, w, ori_h, ori_w, has_cls_token=True):
+    """
+    Calculate absolute positional embeddings. If needed, resize embeddings and remove cls_token
+        dimension for the original embeddings.
+    Args:
+        abs_pos (Tensor): absolute positional embeddings with (1, num_position, C).
+        has_cls_token (bool): If true, has 1 embedding in abs_pos for cls token.
+        hw (Tuple): size of input image tokens.
+
+    Returns:
+        Absolute positional embeddings after processing with shape (1, H, W, C)
+    """
+    cls_token = None
+    B, L, C = abs_pos.shape
+    if has_cls_token:
+        cls_token = abs_pos[:, 0:1]
+        abs_pos = abs_pos[:, 1:]
+
+    if ori_h != h or ori_w != w:
+        new_abs_pos = F.interpolate(
+            abs_pos.reshape(1, ori_h, ori_w, -1).permute(0, 3, 1, 2),
+            size=(h, w),
+            mode="bicubic",
+            align_corners=False,
+        ).permute(0, 2, 3, 1).reshape(B, -1, C)
+
+    else:
+        new_abs_pos = abs_pos
+    
+    if cls_token is not None:
+        new_abs_pos = torch.cat([cls_token, new_abs_pos], dim=1)
+    return new_abs_pos
+
+class DropPath(nn.Module):
+    """Drop paths (Stochastic Depth) per sample  (when applied in main path of residual blocks).
+    """
+    def __init__(self, drop_prob=None):
+        super(DropPath, self).__init__()
+        self.drop_prob = drop_prob
+
+    def forward(self, x):
+        return drop_path(x, self.drop_prob, self.training)
+    
+    def extra_repr(self):
+        return 'p={}'.format(self.drop_prob)
+
+class Mlp(nn.Module):
+    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x
+
+class MoEMlp(nn.Module):
+    def __init__(self, num_expert=1, in_features=1024, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0., part_features=256):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.part_features = part_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features - part_features)
+        self.drop = nn.Dropout(drop)
+        
+        self.num_expert = num_expert
+        experts = []
+
+        for i in range(num_expert):
+            experts.append(
+                        nn.Linear(hidden_features, part_features)
+                        )
+        self.experts = nn.ModuleList(experts)
+
+    def forward(self, x, indices):
+
+        expert_x = torch.zeros_like(x[:, :, -self.part_features:], device=x.device, dtype=x.dtype)
+
+        x = self.fc1(x)
+        x = self.act(x)
+        shared_x = self.fc2(x)
+        indices = indices.view(-1, 1, 1)
+
+        # to support ddp training
+        for i in range(self.num_expert):
+            selectedIndex = (indices == i)
+            current_x = self.experts[i](x) * selectedIndex
+            expert_x = expert_x + current_x
+
+        x = torch.cat([shared_x, expert_x], dim=-1)
+
+        return x
+
+class Attention(nn.Module):
+    def __init__(
+            self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0.,
+            proj_drop=0., attn_head_dim=None,):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.dim = dim
+
+        if attn_head_dim is not None:
+            head_dim = attn_head_dim
+        all_head_dim = head_dim * self.num_heads
+
+        self.scale = qk_scale or head_dim ** -0.5
+
+        self.qkv = nn.Linear(dim, all_head_dim * 3, bias=qkv_bias)
+
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(all_head_dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+    def forward(self, x):
+        B, N, C = x.shape
+        qkv = self.qkv(x)
+        qkv = qkv.reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)
+        q, k, v = qkv[0], qkv[1], qkv[2]   # make torchscript happy (cannot use tensor as tuple)
+
+        q = q * self.scale
+        attn = (q @ k.transpose(-2, -1))
+
+        attn = attn.softmax(dim=-1)
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B, N, -1)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+
+        return x
+
+class Block(nn.Module):
+
+    def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, 
+                 drop=0., attn_drop=0., drop_path=0., act_layer=nn.GELU, 
+                 norm_layer=nn.LayerNorm, attn_head_dim=None, num_expert=1, part_features=None
+                 ):
+        super().__init__()
+
+        self.norm1 = norm_layer(dim)
+        self.attn = Attention(
+            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale,
+            attn_drop=attn_drop, proj_drop=drop, attn_head_dim=attn_head_dim
+            )
+
+        # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = MoEMlp(num_expert=num_expert, in_features=dim, hidden_features=mlp_hidden_dim, 
+                            act_layer=act_layer, drop=drop, part_features=part_features)
+
+    def forward(self, x, indices=None):
+
+        x = x + self.drop_path(self.attn(self.norm1(x)))
+        x = x + self.drop_path(self.mlp(self.norm2(x), indices))
+        return x
+
+
+class PatchEmbed(nn.Module):
+    """ Image to Patch Embedding
+    """
+    def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, ratio=1):
+        super().__init__()
+        img_size = to_2tuple(img_size)
+        patch_size = to_2tuple(patch_size)
+        num_patches = (img_size[1] // patch_size[1]) * (img_size[0] // patch_size[0]) * (ratio ** 2)
+        self.patch_shape = (int(img_size[0] // patch_size[0] * ratio), int(img_size[1] // patch_size[1] * ratio))
+        self.origin_patch_shape = (int(img_size[0] // patch_size[0]), int(img_size[1] // patch_size[1]))
+        self.img_size = img_size
+        self.patch_size = patch_size
+        self.num_patches = num_patches
+
+        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=(patch_size[0] // ratio), padding=4 + 2 * (ratio//2-1))
+
+    def forward(self, x, **kwargs):
+        B, C, H, W = x.shape
+        x = self.proj(x)
+        Hp, Wp = x.shape[2], x.shape[3]
+
+        x = x.flatten(2).transpose(1, 2)
+        return x, (Hp, Wp)
+
+
+class HybridEmbed(nn.Module):
+    """ CNN Feature Map Embedding
+    Extract feature map from CNN, flatten, project to embedding dim.
+    """
+    def __init__(self, backbone, img_size=224, feature_size=None, in_chans=3, embed_dim=768):
+        super().__init__()
+        assert isinstance(backbone, nn.Module)
+        img_size = to_2tuple(img_size)
+        self.img_size = img_size
+        self.backbone = backbone
+        if feature_size is None:
+            with torch.no_grad():
+                training = backbone.training
+                if training:
+                    backbone.eval()
+                o = self.backbone(torch.zeros(1, in_chans, img_size[0], img_size[1]))[-1]
+                feature_size = o.shape[-2:]
+                feature_dim = o.shape[1]
+                backbone.train(training)
+        else:
+            feature_size = to_2tuple(feature_size)
+            feature_dim = self.backbone.feature_info.channels()[-1]
+        self.num_patches = feature_size[0] * feature_size[1]
+        self.proj = nn.Linear(feature_dim, embed_dim)
+
+    def forward(self, x):
+        x = self.backbone(x)[-1]
+        x = x.flatten(2).transpose(1, 2)
+        x = self.proj(x)
+        return x
+
+class ViTMoE(nn.Module):
+    def __init__(self,
+                 img_size=224, patch_size=16, in_chans=3, num_classes=80, embed_dim=768, depth=12,
+                 num_heads=12, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop_rate=0., attn_drop_rate=0.,
+                 drop_path_rate=0., hybrid_backbone=None, norm_layer=None, use_checkpoint=False, 
+                 frozen_stages=-1, ratio=1, last_norm=True,
+                 patch_padding='pad', freeze_attn=False, freeze_ffn=False,
+                 num_expert=1, part_features=None
+                 ):
+        # Protect mutable default arguments
+        super(ViTMoE, self).__init__()
+        norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)
+        self.num_classes = num_classes
+        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
+        self.frozen_stages = frozen_stages
+        self.use_checkpoint = use_checkpoint
+        self.patch_padding = patch_padding
+        self.freeze_attn = freeze_attn
+        self.freeze_ffn = freeze_ffn
+        self.depth = depth
+
+        if hybrid_backbone is not None:
+            self.patch_embed = HybridEmbed(
+                hybrid_backbone, img_size=img_size, in_chans=in_chans, embed_dim=embed_dim)
+        else:
+            self.patch_embed = PatchEmbed(
+                img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim, ratio=ratio)
+        num_patches = self.patch_embed.num_patches
+
+        self.part_features = part_features
+
+        self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))
+
+        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
+
+        self.blocks = nn.ModuleList([
+            Block(
+                dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
+                drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer,
+                num_expert=num_expert, part_features=part_features
+                )
+            for i in range(depth)])
+
+        self.last_norm = norm_layer(embed_dim) if last_norm else nn.Identity()
+
+        if self.pos_embed is not None:
+            trunc_normal_(self.pos_embed, std=.02)
+
+        self._freeze_stages()
+
+    def _freeze_stages(self):
+        """Freeze parameters."""
+        if self.frozen_stages >= 0:
+            self.patch_embed.eval()
+            for param in self.patch_embed.parameters():
+                param.requires_grad = False
+
+        for i in range(1, self.frozen_stages + 1):
+            m = self.blocks[i]
+            m.eval()
+            for param in m.parameters():
+                param.requires_grad = False
+
+        if self.freeze_attn:
+            for i in range(0, self.depth):
+                m = self.blocks[i]
+                m.attn.eval()
+                m.norm1.eval()
+                for param in m.attn.parameters():
+                    param.requires_grad = False
+                for param in m.norm1.parameters():
+                    param.requires_grad = False
+
+        if self.freeze_ffn:
+            self.pos_embed.requires_grad = False
+            self.patch_embed.eval()
+            for param in self.patch_embed.parameters():
+                param.requires_grad = False
+            for i in range(0, self.depth):
+                m = self.blocks[i]
+                m.mlp.eval()
+                m.norm2.eval()
+                for param in m.mlp.parameters():
+                    param.requires_grad = False
+                for param in m.norm2.parameters():
+                    param.requires_grad = False
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super().init_weights(pretrained, patch_padding=self.patch_padding, part_features=self.part_features)
+
+        if pretrained is None:
+            def _init_weights(m):
+                if isinstance(m, nn.Linear):
+                    trunc_normal_(m.weight, std=.02)
+                    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)
+
+            self.apply(_init_weights)
+
+    def get_num_layers(self):
+        return len(self.blocks)
+
+    @torch.jit.ignore
+    def no_weight_decay(self):
+        return {'pos_embed', 'cls_token'}
+
+    def forward_features(self, x, dataset_source=None):
+        B, C, H, W = x.shape
+        x, (Hp, Wp) = self.patch_embed(x)
+
+        if self.pos_embed is not None:
+            # fit for multiple GPU training
+            # since the first element for pos embed (sin-cos manner) is zero, it will cause no difference
+            x = x + self.pos_embed[:, 1:] + self.pos_embed[:, :1]
+
+        for blk in self.blocks:
+            if self.use_checkpoint:
+                x = checkpoint.checkpoint(blk, x, dataset_source)
+            else:
+                x = blk(x, dataset_source)
+
+        x = self.last_norm(x)
+
+        xp = x.permute(0, 2, 1).reshape(B, -1, Hp, Wp).contiguous()
+
+        return xp
+
+    def forward(self, x, dataset_source=None):
+        x = self.forward_features(x, dataset_source)
+        return x
+
+    def train(self, mode=True):
+        """Convert the model into training mode."""
+        super().train(mode)
+        self._freeze_stages()
+
+class Head(nn.Module):
+    def __init__(self, in_channels,
+                 out_channels,
+                 num_deconv_layers=3,
+                 num_deconv_filters=(256, 256, 256),
+                 num_deconv_kernels=(4, 4, 4),):
+        super().__init__()
+        self.in_channels = in_channels
+        self.deconv_layers = self._make_deconv_layer(num_deconv_layers, num_deconv_filters, num_deconv_kernels)
+        self.final_layer = nn.Conv2d(in_channels=num_deconv_filters[-1], out_channels=out_channels, 
+                                      kernel_size=1, stride=1, padding=0)
+
+    def _make_deconv_layer(self, num_layers, num_filters, num_kernels):
+        layers = []
+        for i in range(num_layers):
+            kernel, padding, output_padding = \
+                self._get_deconv_cfg(num_kernels[i])
+
+            planes = num_filters[i]
+            layers.append(
+                nn.ConvTranspose2d(
+                    in_channels=self.in_channels,
+                    out_channels=planes,
+                    kernel_size=kernel,
+                    stride=2,
+                    padding=padding,
+                    output_padding=output_padding,
+                    bias=False))
+            layers.append(nn.BatchNorm2d(planes))
+            layers.append(nn.ReLU(inplace=True))
+            self.in_channels = planes
+
+        return nn.Sequential(*layers)
+    
+    @staticmethod
+    def _get_deconv_cfg(deconv_kernel):
+        """Get configurations for deconv layers."""
+        if deconv_kernel == 4:
+            padding = 1
+            output_padding = 0
+        elif deconv_kernel == 3:
+            padding = 1
+            output_padding = 1
+        elif deconv_kernel == 2:
+            padding = 0
+            output_padding = 0
+        else:
+            raise ValueError(f'Not supported num_kernels ({deconv_kernel}).')
+
+        return deconv_kernel, padding, output_padding
+
+    def forward(self, x):
+        """Forward function."""
+        x = self.deconv_layers(x)
+        x = self.final_layer(x)
+        return x
+
+
+class ComposeVit(nn.Module):
+    def __init__(self):
+        super().__init__()
+        cfg_backbone = dict(
+            img_size=(256, 192),
+            patch_size=16,
+            embed_dim=768,
+            depth=12,
+            num_heads=12,
+            ratio=1,
+            use_checkpoint=False,
+            mlp_ratio=4,
+            qkv_bias=True,
+            drop_path_rate=0.3,
+            num_expert=6,
+            part_features=192
+        )
+        cfg_head = dict(
+            in_channels=768,
+            out_channels=17,
+            num_deconv_layers=2,
+            num_deconv_filters=(256, 256),
+            num_deconv_kernels=(4, 4),
+        )
+        cfg_head_133 = dict(
+            in_channels=768,
+            out_channels=133,
+            num_deconv_layers=2,
+            num_deconv_filters=(256, 256),
+            num_deconv_kernels=(4, 4),
+        )
+        self.backbone = ViTMoE(**cfg_backbone)
+        self.keypoint_head = Head(**cfg_head)
+        self.associate_head = Head(**cfg_head_133)
+    
+    def forward(self, x):
+        indices = torch.zeros((x.shape[0]), dtype=torch.long, device=x.device)
+        back_out = self.backbone(x, indices)
+        out = self.keypoint_head(back_out)
+        if True:
+            indices += 5 # 最后一个是whole body dataset
+            back_133 = self.backbone(x, indices)
+            out_133 = self.associate_head(back_133)
+            out_foot = out_133[:, 17:23]
+            out = torch.cat([out, out_foot], dim=1)
+            if False:
+                import cv2
+                vis = x[0].permute(1, 2, 0).cpu().numpy()
+                mean= np.array([0.485, 0.456, 0.406]).reshape(1, 1, 3)
+                std=np.array([0.229, 0.224, 0.225]).reshape(1, 1 ,3)
+                vis = np.clip(vis * std + mean, 0., 1.)
+                vis = (vis[:,:,::-1] * 255).astype(np.uint8)
+                value = out_133[0].detach().cpu().numpy()
+                vis_all = []
+                for i in range(value.shape[0]):
+                    _val = np.clip(value[i], 0., 1.)
+                    _val = (_val * 255).astype(np.uint8)
+                    _val = cv2.resize(_val, None, fx=4, fy=4)
+                    _val = cv2.applyColorMap(_val, cv2.COLORMAP_JET)
+                    _vis = cv2.addWeighted(vis, 0.5, _val, 0.5, 0)
+                    vis_all.append(_vis)
+                from easymocap.mytools.vis_base import merge
+                cv2.imwrite('debug.jpg', merge(vis_all))
+
+                import ipdb; ipdb.set_trace()
+        return {
+            'output': out
+        }
+
+from ..basetopdown import BaseTopDownModelCache
+from ..topdown_keypoints import BaseKeypoints
+
+class MyViT(BaseTopDownModelCache, BaseKeypoints):
+    def __init__(self, ckpt='data/models/vitpose+_base.pth', single_person=True, url='https://1drv.ms/u/s!AimBgYV7JjTlgcckRZk1bIAuRa_E1w?e=ylDB2G', **kwargs):
+        super().__init__(name='myvit', bbox_scale=1.25,
+                         res_input=[192, 256], **kwargs)
+        self.single_person = single_person
+        model = ComposeVit()
+        if not os.path.exists(ckpt):
+            print('')
+            print('{} not exists, please download it from {} and place it to {}'.format(ckpt, url, ckpt))
+            print('')
+            raise FileNotFoundError
+        ckpt = torch.load(ckpt, map_location='cpu')['state_dict']
+        ckpt_backbone = {key:val for key, val in ckpt.items() if key.startswith('backbone.')}
+        ckpt_head = {key:val for key, val in ckpt.items() if key.startswith('keypoint_head.')}
+        key_whole = 'associate_keypoint_heads.4.'
+        ckpt_head_133 = {key.replace(key_whole, 'associate_head.'):val for key, val in ckpt.items() if key.startswith(key_whole)}
+        ckpt_backbone.update(ckpt_head)
+        ckpt_backbone.update(ckpt_head_133)
+        state_dict = ckpt_backbone
+        self.load_checkpoint(model, state_dict, prefix='', strict=True)
+        model.eval()
+        self.model = model
+        self.device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
+        self.model.to(self.device)
+
+    def dump(self, cachename, output):
+        _output = output['output']
+        kpts = self.get_max_preds(_output)
+        kpts_ori = self.batch_affine_transform(kpts, output['inv_trans'])
+        kpts = np.concatenate([kpts_ori, kpts[..., -1:]], axis=-1)
+        output = {'keypoints': kpts}
+        super().dump(cachename, output)
+        return output
+
+    def estimate_keypoints(self, bbox, images, imgnames):
+        squeeze = False
+        if not isinstance(images, list):
+            images = [images]
+            imgnames = [imgnames]
+            bbox = [bbox]
+            squeeze = True
+        nViews = len(images)
+        kpts_all = []
+        for nv in range(nViews):
+            _bbox = bbox[nv]
+            if _bbox.shape[0] == 0:
+                if self.single_person:
+                    kpts = np.zeros((1, self.num_joints, 3))
+                else:
+                    kpts = np.zeros((_bbox.shape[0], self.num_joints, 3))
+            else:
+                img = images[nv]
+                # TODO: add flip test
+                out = super().__call__(_bbox, img, imgnames[nv])
+                kpts = out['params']['keypoints']
+            if kpts.shape[-2] == 23:
+                kpts = self.coco23tobody25(kpts)
+            elif kpts.shape[-2] == 17:
+                kpts = self.coco17tobody25(kpts)
+            else:
+                raise NotImplementedError
+            kpts_all.append(kpts)
+        if self.single_person:
+            kpts_all = [k[0] for k in kpts_all]
+            kpts_all = np.stack(kpts_all)
+        if squeeze:
+            kpts_all = kpts_all[0]
+        return {
+            'keypoints': kpts_all
+        }
+
+    def __call__(self, bbox, images, imgnames):
+        return self.estimate_keypoints(bbox, images, imgnames)
+
+if __name__ == '__main__':
+    # Load checkpoint
+    rand_input = torch.rand(1, 3, 256, 192)
+    model = MyViT()

myeasymocap/backbone/yolo/yolo.py

@@ -145,6 +145,21 @@ class YoloWithTrack(BaseYOLOv5):
         self.track_cache[sub]['bbox'].append(select)
         return select
 
+class MultiPerson(BaseYOLOv5):
+    def __init__(self, min_length, max_length, **kwargs):
+        super().__init__(**kwargs)
+        self.min_length = min_length
+        self.max_length = max_length
+        print('[{}] Only keep the bbox in [{}, {}]'.format(self.__class__.__name__, min_length, max_length))
+
+    def select_bbox(self, select, imgname):
+        if select.shape[0] == 0:
+            return select
+        # 判断一下面积
+        area = np.sqrt((select[:, 2] - select[:, 0])*(select[:, 3]-select[:, 1]))
+        valid = (area > self.min_length) & (area < self.max_length)
+        return select[valid]
+
 class DetectToPelvis:
     def __init__(self, key) -> None:
         self.key = key