COCO Model mAP Eval

samples/test/eval_mAP_on_coco_val/coco_eval.py

@@ -5,7 +5,7 @@ import os
 if __name__ == '__main__':
   path = os.path.abspath(os.path.join(os.getcwd(),"../../.."))
   pred_json = 'pd_coco.json'
-    anno_json = path + '/val2017/gt_coco.json'
+  anno_json = path + '/val2017/instances_val2017.json'
 
   # use COCO API to load forecast results and annotations
   cocoGt = COCO(anno_json)

samples/test/eval_mAP_on_coco_val/eval_mAP_on_coco_val.cpp

@@ -65,7 +65,7 @@ int main(int argc, char *argv[])
 
     //create pred file
     std::string val_image_name = GetImageFileName(val_image[i]);
-    std::string filename = folder+ val_image_name + ".txt";
+    std::string filename = folder+"/"+ val_image_name + ".txt";
     std::ofstream file(filename);
     file.is_open();
     file<<std::fixed<<std::setprecision(6);

samples/test/eval_mAP_on_coco_val/gt2cocojson.py

@@ -1,123 +0,0 @@
-# -*- coding: UTF-8 -*-
-"""
-  @Author: lxm
-  
-  @version V1.0
-"""
-import datetime
-import json
-import os
-import cv2
-
-# 将yolo格式的数据集转换成coco格式的数据集
-
-# 读取文件夹下的所有文件
-path = os.path.abspath(os.path.join(os.getcwd(),"../../.."))
-images_path = path+'/val2017/val2017'
-labels_path = path+'/val2017/labels'
-
-coco_json_save = path + '/val2017/gt_coco.json'
-
-
-# 创建coco格式的json文件
-coco_json = {
-	'info': {
-		"description": "COCOVal Dataset",
-		"url": "www.amov.com",
-		"version": "1.0",
-		"year": 2023,
-		"contributor": "lxm",
-		"date_created": datetime.datetime.utcnow().isoformat(' ')
-	},
-	"licenses": [
-		{
-			"url": "http://creativecommons.org/licenses/by-nc-sa/2.0/",
-			"id": 1,
-			"name": "Attribution-NonCommercial-ShareAlike License"
-		}
-	],
-	'images': [],
-	'annotations': [],
-	'categories': []
-}
-
-
-# COCO classes
-classes = ["person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck",
-    "boat", "traffic light", "fire hydrant", "stop sign", "parking meter", "bench",
-    "bird", "cat", "dog", "horse", "sheep", "cow", "elephant", "bear", "zebra",
-    "giraffe", "backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee",
-    "skis", "snowboard", "sports ball", "kite", "baseball bat", "baseball glove",
-    "skateboard", "surfboard", "tennis racket", "bottle", "wine glass", "cup", "fork",
-    "knife", "spoon", "bowl", "banana", "apple", "sandwich", "orange", "broccoli",
-    "carrot", "hot dog", "pizza", "donut", "cake", "chair", "couch", "potted plant",
-    "bed", "dining table", "toilet", "tv", "laptop", "mouse", "remote", "keyboard",
-    "cell phone", "microwave", "oven", "toaster", "sink", "refrigerator", "book",
-    "clock", "vase", "scissors", "teddy bear", "hair drier", "toothbrush"]
-
-# 创建coco格式的json文件
-for i, c in enumerate(classes):
-	coco_json['categories'].append({'id': i + 1, 'name': c, 'supercategory': c})
-
-
-
-
-
-
-# 读取images文件夹下的所有文件
-images = os.listdir(images_path)
-for image in images:
-	# 获取图片名和后缀
-	image_name, image_suffix = os.path.splitext(image)
-	# 获取图片的宽和高
-	image_path = images_path + '/' + image
-	img = cv2.imread(image_path)
-	height, width, _ = img.shape
-	# 添加图片信息
-	coco_json['images'].append({
-		'id': int(image_name),
-		'file_name': image,
-		'width': width,
-		'height': height,
-		'date_captured': datetime.datetime.utcnow().isoformat(' '),
-		'license': 1
-	})
-	
-	# 读取图片对应的标签文件
-	label_path = labels_path + '/' + image_name + '.txt'
-	if not os.path.exists(label_path):
-		continue
-	with open(label_path, 'r') as f:
-		labels = f.readlines()
-		labels = [l.strip() for l in labels]
-		for j, label in enumerate(labels):
-			label = label.split(' ')
-			# 获取类别id
-			category_id = int(label[0])
-			# 将yolo格式的数据转换成coco格式的数据
-			x = float(label[1]) * width
-			y = float(label[2]) * height
-			w = float(label[3]) * width
-			h = float(label[4]) * height
-			xmin = x - w / 2
-			ymin = y - h / 2
-			xmax = x + w / 2
-			ymax = y + h / 2
-			# 添加bbox信息
-			# 添加bbox信息
-			coco_json['annotations'].append({
-				'image_id': int(image_name),
-				'category_id': category_id + 1,
-				'bbox': [xmin, ymin, w, h],
-				'id': len(coco_json['annotations']),
-				'area': w * h,
-				'iscrowd': 0,
-				'segmentation': [],
-				'attributes': ""
-			})
-
-# 保存json文件
-with open(coco_json_save, 'w') as f:
-	json.dump(coco_json, f, indent=2)
-
-print(len(coco_json['images']), len(coco_json['annotations']), len(coco_json['categories']), 'Done!')

samples/test/eval_mAP_on_coco_val/pd2cocojson.py

@@ -14,6 +14,10 @@ coco_json_save ='pd_coco.json'
 # config coco_json
 coco_json = []
 
+# remap the id of the coco dataset
+id_map = {1: 0, 2: 1, 3: 2, 4: 3, 5: 4, 6: 5, 7: 6, 8: 7, 9: 8, 10: 9, 11: 10, 13: 11, 14: 12, 15: 13, 16: 14, 17: 15, 18: 16, 19: 17, 20: 18, 21: 19, 22: 20, 23: 21, 24: 22, 25: 23, 27: 24, 28: 25, 31: 26, 32: 27, 33: 28, 34: 29, 35: 30, 36: 31, 37: 32, 38: 33, 39: 34, 40: 35, 41: 36, 42: 37, 43: 38, 44: 39, 46: 40, 47: 41, 48: 42, 49: 43, 50: 44, 51: 45, 52: 46, 53: 47, 54: 48, 55: 49, 56: 50, 57: 51, 58: 52, 59: 53, 60: 54, 61: 55, 62: 56, 63: 57, 64: 58, 65: 59, 67: 60, 70: 61, 72: 62, 73: 63, 74: 64, 75: 65, 76: 66, 77: 67, 78: 68, 79: 69, 80: 70, 81: 71, 82: 72, 84: 73, 85: 74, 86: 75, 87: 76, 88: 77, 89: 78, 90: 79}
+reid_mp = { value: key for key, value in id_map.items()}
+
 # load images dir
 images = os.listdir(images_path)
 for image in images:
@@ -46,7 +50,7 @@ for image in images:
 			
       coco_json.append({
         'image_id': int(image_name),
-            'category_id': category_id + 1,
+        'category_id': int(reid_mp[category_id]),
         'bbox': [xmin, ymin, w, h],
         'score': float(pred[5]),
         'area': w * h})