File size: 9,891 Bytes
2214a66 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 | """
COCO Patch-based Evaluation
直接输入超分后的patches路径，评估Object Detection和Instance Segmentation指标
Metrics:
- Object Detection: AP^b, AP^b_50, AP^b_75
- Instance Segmentation: AP^m, AP^m_50, AP^m_75
"""
import os
import json
from pathlib import Path
from typing import Dict, List, Tuple
import torch
from torch.utils.data import DataLoader, Dataset
from torchvision.models.detection import (
maskrcnn_resnet50_fpn_v2,
MaskRCNN_ResNet50_FPN_V2_Weights,
)
from torchvision.transforms import functional as TF
from PIL import Image
import numpy as np
from tqdm import tqdm
from pycocotools.coco import COCO
from pycocotools.cocoeval import COCOeval
from pycocotools import mask as mask_util
# ============================================================================
# 配置参数 - 请修改这里
# ============================================================================
CONFIG = {
# SR patches 目录 (你的超分模型输出，512x512)
# 图片命名需与 prepare_coco_patches.py 生成的一致
'sr_dir': '/home/wanghongbo06/baipurui/results/COCO/DreamClear/results/output',
# Patch annotations 文件 (prepare_coco_patches.py 生成的)
'ann_file': '/home/wanghongbo06/baipurui/DATA/COCO_patch/patch_annotations.json',
# 推理配置
'device': 'cuda',
'batch_size': 8,
'num_workers': 4,
# 输出
'output': './coco_patch_eval_results.json',
}
# ============================================================================
# COCO类别ID映射
COCO_CATEGORY_IDS = [
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17,
18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34,
35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 46, 47, 48, 49,
50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63,
64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81,
82, 84, 85, 86, 87, 88, 89, 90
]
class PatchDataset(Dataset):
"""Patch数据集"""
def __init__(self, image_dir: str, ann_file: str):
self.image_dir = Path(image_dir)
self.coco = COCO(ann_file)
# 过滤出存在的图片
all_image_ids = list(self.coco.imgs.keys())
self.image_ids = []
missing_count = 0
print("Checking image files...")
for img_id in all_image_ids:
img_info = self.coco.imgs[img_id]
img_path = self.image_dir / img_info['file_name']
if img_path.exists():
self.image_ids.append(img_id)
else:
missing_count += 1
if missing_count > 0:
print(f"Warning: {missing_count} images not found, skipped.")
print(f"Valid images: {len(self.image_ids)} / {len(all_image_ids)}")
def __len__(self):
return len(self.image_ids)
def __getitem__(self, idx: int) -> Dict:
image_id = self.image_ids[idx]
img_info = self.coco.imgs[image_id]
img_path = self.image_dir / img_info['file_name']
image = Image.open(img_path).convert('RGB')
image_tensor = TF.to_tensor(image)
return {
'image': image_tensor,
'image_id': image_id,
}
def collate_fn(batch):
return batch
class PatchEvaluator:
"""Patch评估器"""
def __init__(self, ann_file: str, device: str = 'cuda'):
self.ann_file = ann_file
self.device = device
self.coco_gt = COCO(ann_file)
print("Loading Mask R-CNN ResNet50 FPN v2...")
weights = MaskRCNN_ResNet50_FPN_V2_Weights.DEFAULT
self.detector = maskrcnn_resnet50_fpn_v2(weights=weights)
self.detector.eval()
self.detector.to(device)
# 注意：torchvision Mask R-CNN 输出的 labels 已经是 COCO category_id
# 不需要额外映射！
@torch.no_grad()
def predict(self, images: List[torch.Tensor]) -> List[Dict]:
images = [img.to(self.device) for img in images]
return self.detector(images)
def convert_to_coco_format(self, predictions: List[Dict], image_ids: List[int]) -> Tuple:
bbox_results = []
segm_results = []
for pred, image_id in zip(predictions, image_ids):
boxes = pred['boxes'].cpu().numpy()
labels = pred['labels'].cpu().numpy()
scores = pred['scores'].cpu().numpy()
masks = pred['masks'].cpu().numpy()
for i in range(len(boxes)):
if scores[i] < 0.05: # 提高阈值
continue
x1, y1, x2, y2 = boxes[i]
bbox = [float(x1), float(y1), float(x2 - x1), float(y2 - y1)]
category_id = int(labels[i]) # 直接使用，已经是 COCO category_id
bbox_results.append({
'image_id': int(image_id),
'category_id': category_id,
'bbox': bbox,
'score': float(scores[i]),
})
mask = masks[i, 0]
mask_binary = (mask > 0.5).astype(np.uint8)
rle = mask_util.encode(np.asfortranarray(mask_binary))
rle['counts'] = rle['counts'].decode('utf-8')
segm_results.append({
'image_id': int(image_id),
'category_id': category_id,
'segmentation': rle,
'score': float(scores[i]),
})
return bbox_results, segm_results
def evaluate(self, dataset: PatchDataset, batch_size: int = 8, num_workers: int = 4) -> Dict:
dataloader = DataLoader(
dataset, batch_size=batch_size, shuffle=False,
num_workers=num_workers, collate_fn=collate_fn,
)
all_bbox_results = []
all_segm_results = []
print("Running inference...")
for batch in tqdm(dataloader):
images = [item['image'] for item in batch]
image_ids = [item['image_id'] for item in batch]
predictions = self.predict(images)
bbox_results, segm_results = self.convert_to_coco_format(predictions, image_ids)
all_bbox_results.extend(bbox_results)
all_segm_results.extend(segm_results)
print("\n" + "="*60)
print("Object Detection (Bounding Box) Results:")
print("="*60)
bbox_metrics = self._run_eval(all_bbox_results, 'bbox')
print("\n" + "="*60)
print("Instance Segmentation (Mask) Results:")
print("="*60)
segm_metrics = self._run_eval(all_segm_results, 'segm')
return {
'AP_bbox': bbox_metrics['AP'],
'AP_bbox_50': bbox_metrics['AP_50'],
'AP_bbox_75': bbox_metrics['AP_75'],
'AP_mask': segm_metrics['AP'],
'AP_mask_50': segm_metrics['AP_50'],
'AP_mask_75': segm_metrics['AP_75'],
}
def _run_eval(self, results: List[Dict], iou_type: str) -> Dict:
if not results:
print(f"No predictions for {iou_type}!")
return {'AP': 0.0, 'AP_50': 0.0, 'AP_75': 0.0}
try:
coco_dt = self.coco_gt.loadRes(results)
coco_eval = COCOeval(self.coco_gt, coco_dt, iou_type)
coco_eval.evaluate()
coco_eval.accumulate()
coco_eval.summarize()
return {
'AP': coco_eval.stats[0],
'AP_50': coco_eval.stats[1],
'AP_75': coco_eval.stats[2],
}
except KeyError as e:
print(f"Error evaluating {iou_type}: {e}")
print(f"Some GT annotations may be missing 'segmentation' field.")
print(f"Skipping {iou_type} evaluation.")
return {'AP': 0.0, 'AP_50': 0.0, 'AP_75': 0.0}
def main():
device = CONFIG['device']
if device == 'cuda' and not torch.cuda.is_available():
print("CUDA not available, using CPU")
device = 'cpu'
# 自动生成 output 路径：根据 sr_dir 最后一个目录名
sr_dir = Path(CONFIG['sr_dir'])
baseline_name = sr_dir.name # 获取最后一个目录名，如 'sr', 'gt', 'bicubic' 等
output_path = Path(f"./coco_patch_eval_results_{baseline_name}.json")
print(f"Loading SR patches from: {CONFIG['sr_dir']}")
print(f"Loading annotations from: {CONFIG['ann_file']}")
print(f"Output will be saved to: {output_path}")
dataset = PatchDataset(
image_dir=CONFIG['sr_dir'],
ann_file=CONFIG['ann_file'],
)
print(f"Dataset size: {len(dataset)} patches")
evaluator = PatchEvaluator(ann_file=CONFIG['ann_file'], device=device)
print("\nStarting evaluation...")
results = evaluator.evaluate(
dataset=dataset,
batch_size=CONFIG['batch_size'],
num_workers=CONFIG['num_workers'],
)
print("\n" + "="*60)
print("EVALUATION SUMMARY")
print("="*60)
print("\nObject Detection (Bounding Box):")
print(f" AP^b : {results['AP_bbox']*100:.2f}")
print(f" AP^b_50 : {results['AP_bbox_50']*100:.2f}")
print(f" AP^b_75 : {results['AP_bbox_75']*100:.2f}")
print("\nInstance Segmentation (Mask):")
print(f" AP^m : {results['AP_mask']*100:.2f}")
print(f" AP^m_50 : {results['AP_mask_50']*100:.2f}")
print(f" AP^m_75 : {results['AP_mask_75']*100:.2f}")
output_path.parent.mkdir(parents=True, exist_ok=True)
with open(output_path, 'w') as f:
json.dump(results, f, indent=2)
print(f"\nResults saved to {output_path}")
if __name__ == '__main__':
main()
|