Remove duplicate line in setup.cfg (ultralytics#9380)

data/dt.yaml

@@ -0,0 +1,27 @@
+# YOLOv5 🚀 by Ultralytics, GPL-3.0 license
+# COCO 2017 dataset http://cocodataset.org by Microsoft
+# Example usage: python train.py --data coco.yaml
+# parent
+# ├── yolov5
+# └── datasets
+#     └── coco  ← downloads here (20.1 GB)
+
+
+# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
+path: D:\detection\datasets\union2voc_multiClass\yolo_dataset  # dataset root dir
+train: train.txt  # train images (relative to 'path') 118287 images
+val: val.txt  # val images (relative to 'path') 5000 images
+test: test.txt  # 20288 of 40670 images, submit to https://competitions.codalab.org/competitions/20794
+
+# Classes
+names:
+  0: Car
+  1: Bus
+  2: Cyclist
+  3: Pedestrian
+  4: driverless_car
+  5: Truck
+  6: Tricyclist
+  7: Trafficcone
+
+# Download script/URL (optional)

utils/dataloaders.py

@@ -114,7 +114,8 @@ def create_dataloader(path,
                       image_weights=False,
                       quad=False,
                       prefix='',
-                      shuffle=False):
+                      shuffle=False,
+                      size_conf=None):
     if rect and shuffle:
         LOGGER.warning('WARNING: --rect is incompatible with DataLoader shuffle, setting shuffle=False')
         shuffle = False
@@ -131,7 +132,8 @@ def create_dataloader(path,
             stride=int(stride),
             pad=pad,
             image_weights=image_weights,
-            prefix=prefix)
+            prefix=prefix,
+            size_conf=size_conf)
 
     batch_size = min(batch_size, len(dataset))
     nd = torch.cuda.device_count()  # number of CUDA devices
@@ -393,7 +395,8 @@ def __init__(self,
                  single_cls=False,
                  stride=32,
                  pad=0.0,
-                 prefix=''):
+                 prefix='',
+                 size_conf=None):
         self.img_size = img_size
         self.augment = augment
         self.hyp = hyp
@@ -430,11 +433,12 @@ def __init__(self,
         self.label_files = img2label_paths(self.im_files)  # labels
         cache_path = (p if p.is_file() else Path(self.label_files[0]).parent).with_suffix('.cache')
         try:
+            cache_path.unlink()  # remove old cache
             cache, exists = np.load(cache_path, allow_pickle=True).item(), True  # load dict
             assert cache['version'] == self.cache_version  # matches current version
             assert cache['hash'] == get_hash(self.label_files + self.im_files)  # identical hash
         except Exception:
-            cache, exists = self.cache_labels(cache_path, prefix), False  # run cache ops
+            cache, exists = self.cache_labels(cache_path, prefix, size_conf), False  # run cache ops
 
         # Display cache
         nf, nm, ne, nc, n = cache.pop('results')  # found, missing, empty, corrupt, total
@@ -517,13 +521,14 @@ def __init__(self,
                 pbar.desc = f'{prefix}Caching images ({gb / 1E9:.1f}GB {cache_images})'
             pbar.close()
 
-    def cache_labels(self, path=Path('./labels.cache'), prefix=''):
+    def cache_labels(self, path=Path('./labels.cache'), prefix='', size_conf=None):
         # Cache dataset labels, check images and read shapes
         x = {}  # dict
         nm, nf, ne, nc, msgs = 0, 0, 0, 0, []  # number missing, found, empty, corrupt, messages
         desc = f"{prefix}Scanning '{path.parent / path.stem}' images and labels..."
         with Pool(NUM_THREADS) as pool:
-            pbar = tqdm(pool.imap(verify_image_label, zip(self.im_files, self.label_files, repeat(prefix))),
+            pbar = tqdm(pool.imap(verify_image_label,
+                                  zip(self.im_files, self.label_files, repeat(prefix), repeat(size_conf))),
                         desc=desc,
                         total=len(self.im_files),
                         bar_format=BAR_FORMAT)
@@ -902,7 +907,7 @@ def autosplit(path=DATASETS_DIR / 'coco128/images', weights=(0.9, 0.1, 0.0), ann
 
 def verify_image_label(args):
     # Verify one image-label pair
-    im_file, lb_file, prefix = args
+    im_file, lb_file, prefix, size_conf = args
     nm, nf, ne, nc, msg, segments = 0, 0, 0, 0, '', []  # number (missing, found, empty, corrupt), message, segments
     try:
         # verify images
@@ -928,6 +933,14 @@ def verify_image_label(args):
                     segments = [np.array(x[1:], dtype=np.float32).reshape(-1, 2) for x in lb]  # (cls, xy1...)
                     lb = np.concatenate((classes.reshape(-1, 1), segments2boxes(segments)), 1)  # (cls, xywh)
                 lb = np.array(lb, dtype=np.float32)
+                if size_conf is not None:
+                    size_thres = np.array([size_conf[int(i)] for i in lb[:, 0]], dtype=np.float32) ** 2
+                    areas = (lb[:, 3:] * np.array(shape, dtype=np.float32)).prod(1)
+                    idx = (areas > size_thres[:, 0]) & (areas <= size_thres[:, 1])
+                    if idx.any():
+                        lb = lb[idx.T]
+                    else:
+                        lb = np.zeros((0, 5), dtype=np.float32)
             nl = len(lb)
             if nl:
                 assert lb.shape[1] == 5, f'labels require 5 columns, {lb.shape[1]} columns detected'

utils/metrics.py

@@ -91,8 +91,9 @@ def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='.', names
     i = smooth(f1.mean(0), 0.1).argmax()  # max F1 index
     p, r, f1 = p[:, i], r[:, i], f1[:, i]
     tp = (r * nt).round()  # true positives
+    fn = nt - tp
     fp = (tp / (p + eps) - tp).round()  # false positives
-    return tp, fp, p, r, f1, ap, unique_classes.astype(int)
+    return tp, fp, fn, p, r, f1, ap, unique_classes.astype(int)
 
 
 def compute_ap(recall, precision):
@@ -177,14 +178,14 @@ def process_batch(self, detections, labels):
                 if not any(m1 == i):
                     self.matrix[dc, self.nc] += 1  # background FN
 
-    def matrix(self):
+    def get_matrix(self):
         return self.matrix
 
     def tp_fp(self):
         tp = self.matrix.diagonal()  # true positives
         fp = self.matrix.sum(1) - tp  # false positives
-        # fn = self.matrix.sum(0) - tp  # false negatives (missed detections)
-        return tp[:-1], fp[:-1]  # remove background class
+        fn = self.matrix.sum(0) - tp  # false negatives (missed detections)
+        return tp[:-1], fp[:-1], fn[:-1]  # remove background class
 
     @TryExcept('WARNING: ConfusionMatrix plot failure: ')
     def plot(self, normalize=True, save_dir='', names=()):

val.py

@@ -265,7 +265,7 @@ def run(
     # Compute metrics
     stats = [torch.cat(x, 0).cpu().numpy() for x in zip(*stats)]  # to numpy
     if len(stats) and stats[0].any():
-        tp, fp, p, r, f1, ap, ap_class = ap_per_class(*stats, plot=plots, save_dir=save_dir, names=names)
+        tp, fp, fn, p, r, f1, ap, ap_class = ap_per_class(*stats, plot=plots, save_dir=save_dir, names=names)
         ap50, ap = ap[:, 0], ap.mean(1)  # AP@0.5, AP@0.5:0.95
         mp, mr, map50, map = p.mean(), r.mean(), ap50.mean(), ap.mean()
     nt = np.bincount(stats[3].astype(int), minlength=nc)  # number of targets per class