Robust `scipy.cluster.vq.kmeans` too few points (#6668)

utils/autoanchor.py

 # YOLOv5 🚀 by Ultralytics, GPL-3.0 license
 """
-Auto-anchor utils
+AutoAnchor utils
 """
 
 import random
@@ -81,6 +81,7 @@ def kmean_anchors(dataset='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen
     """
     from scipy.cluster.vq import kmeans
 
+    npr = np.random
     thr = 1 / thr
 
     def metric(k, wh):  # compute metrics
@@ -121,14 +122,15 @@ def kmean_anchors(dataset='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen
     if i:
         LOGGER.info(f'{PREFIX}WARNING: Extremely small objects found. {i} of {len(wh0)} labels are < 3 pixels in size.')
     wh = wh0[(wh0 >= 2.0).any(1)]  # filter > 2 pixels
-    # wh = wh * (np.random.rand(wh.shape[0], 1) * 0.9 + 0.1)  # multiply by random scale 0-1
+    # wh = wh * (npr.rand(wh.shape[0], 1) * 0.9 + 0.1)  # multiply by random scale 0-1
 
     # Kmeans calculation
     LOGGER.info(f'{PREFIX}Running kmeans for {n} anchors on {len(wh)} points...')
     s = wh.std(0)  # sigmas for whitening
-    k, dist = kmeans(wh / s, n, iter=30)  # points, mean distance
-    assert len(k) == n, f'{PREFIX}ERROR: scipy.cluster.vq.kmeans requested {n} points but returned only {len(k)}'
-    k *= s
+    k = kmeans(wh / s, n, iter=30)[0] * s  # points
+    if len(k) != n:  # kmeans may return fewer points than requested if wh is insufficient or too similar
+        LOGGER.warning(f'{PREFIX}WARNING: scipy.cluster.vq.kmeans returned only {len(k)} of {n} requested points')
+        k = np.sort(npr.rand(n * 2)).reshape(n, 2) * img_size  # random init
     wh = torch.tensor(wh, dtype=torch.float32)  # filtered
     wh0 = torch.tensor(wh0, dtype=torch.float32)  # unfiltered
     k = print_results(k, verbose=False)
@@ -146,7 +148,6 @@ def kmean_anchors(dataset='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen
     # fig.savefig('wh.png', dpi=200)
 
     # Evolve
-    npr = np.random
     f, sh, mp, s = anchor_fitness(k), k.shape, 0.9, 0.1  # fitness, generations, mutation prob, sigma
     pbar = tqdm(range(gen), desc=f'{PREFIX}Evolving anchors with Genetic Algorithm:')  # progress bar
     for _ in pbar: