59 lines
1.7 KiB
Python
59 lines
1.7 KiB
Python
import json
|
||
import math
|
||
import cv2
|
||
import numpy as np
|
||
import torch
|
||
from collections import namedtuple
|
||
from typing import List, Union, NamedTuple
|
||
from .dwpose import *
|
||
|
||
def numpy2torch(np_image: np.ndarray) -> torch.Tensor:
|
||
""" [H, W, C] => [B=1, H, W, C]"""
|
||
return torch.from_numpy(np_image.astype(np.float32) / 255).unsqueeze(0)
|
||
|
||
class RenderKps:
|
||
@classmethod
|
||
def INPUT_TYPES(s):
|
||
return {
|
||
"required": {
|
||
"kps": ("POSE_KEYPOINT",),
|
||
"render_body": ("BOOLEAN", {"default": True}),
|
||
}
|
||
}
|
||
|
||
RETURN_TYPES = ("IMAGE",)
|
||
FUNCTION = "render"
|
||
CATEGORY = "Snap Processing"
|
||
|
||
def render(self, kps, render_body):
|
||
frames = kps if isinstance(kps, list) else [kps]
|
||
|
||
imgs = []
|
||
for data in frames:
|
||
if isinstance(data, list):
|
||
data = data[0]
|
||
if isinstance(data, str):
|
||
try:
|
||
data = json.loads(data)
|
||
except json.JSONDecodeError:
|
||
continue
|
||
if not isinstance(data, dict):
|
||
continue
|
||
|
||
poses, _, H, W = decode_json_as_poses(data)
|
||
np_img = draw_poses(poses, H, W, render_body)
|
||
imgs.append(torch.from_numpy(np_img.astype(np.float32) / 255.0)) # [H,W,3]
|
||
|
||
if not imgs:
|
||
imgs = [torch.zeros((1, 1, 3), dtype=torch.float32)]
|
||
|
||
# ── 把 [H,W,3] 堆到一起 → [B,H,W,3]
|
||
batch = torch.stack(imgs, dim=0) # B × H × W × 3
|
||
print("batch shape:", batch.shape) # e.g. torch.Size([2,768,512,3])
|
||
|
||
# 返回一个 tuple，符合 RETURN_TYPES
|
||
return (batch,)
|
||
|
||
NODE_CLASS_MAPPINGS = {
|
||
"RenderKps": RenderKps,
|
||
} |