|
|
@@ -6,6 +6,8 @@ import yaml |
|
|
|
from scipy.cluster.vq import kmeans |
|
|
|
from tqdm import tqdm |
|
|
|
|
|
|
|
from utils.general import colorstr |
|
|
|
|
|
|
|
|
|
|
|
def check_anchor_order(m): |
|
|
|
# Check anchor order against stride order for YOLOv5 Detect() module m, and correct if necessary |
|
|
@@ -20,7 +22,8 @@ def check_anchor_order(m): |
|
|
|
|
|
|
|
def check_anchors(dataset, model, thr=4.0, imgsz=640): |
|
|
|
# Check anchor fit to data, recompute if necessary |
|
|
|
print('\nAnalyzing anchors... ', end='') |
|
|
|
prefix = colorstr('blue', 'bold', 'autoanchor') + ': ' |
|
|
|
print(f'\n{prefix}Analyzing anchors... ', end='') |
|
|
|
m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1] # Detect() |
|
|
|
shapes = imgsz * dataset.shapes / dataset.shapes.max(1, keepdims=True) |
|
|
|
scale = np.random.uniform(0.9, 1.1, size=(shapes.shape[0], 1)) # augment scale |
|
|
@@ -35,7 +38,7 @@ def check_anchors(dataset, model, thr=4.0, imgsz=640): |
|
|
|
return bpr, aat |
|
|
|
|
|
|
|
bpr, aat = metric(m.anchor_grid.clone().cpu().view(-1, 2)) |
|
|
|
print('anchors/target = %.2f, Best Possible Recall (BPR) = %.4f' % (aat, bpr), end='') |
|
|
|
print(f'anchors/target = {aat:.2f}, Best Possible Recall (BPR) = {bpr:.4f}', end='') |
|
|
|
if bpr < 0.98: # threshold to recompute |
|
|
|
print('. Attempting to improve anchors, please wait...') |
|
|
|
na = m.anchor_grid.numel() // 2 # number of anchors |
|
|
@@ -46,9 +49,9 @@ def check_anchors(dataset, model, thr=4.0, imgsz=640): |
|
|
|
m.anchor_grid[:] = new_anchors.clone().view_as(m.anchor_grid) # for inference |
|
|
|
m.anchors[:] = new_anchors.clone().view_as(m.anchors) / m.stride.to(m.anchors.device).view(-1, 1, 1) # loss |
|
|
|
check_anchor_order(m) |
|
|
|
print('New anchors saved to model. Update model *.yaml to use these anchors in the future.') |
|
|
|
print(f'{prefix}New anchors saved to model. Update model *.yaml to use these anchors in the future.') |
|
|
|
else: |
|
|
|
print('Original anchors better than new anchors. Proceeding with original anchors.') |
|
|
|
print(f'{prefix}Original anchors better than new anchors. Proceeding with original anchors.') |
|
|
|
print('') # newline |
|
|
|
|
|
|
|
|
|
|
@@ -70,6 +73,7 @@ def kmean_anchors(path='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=10 |
|
|
|
from utils.autoanchor import *; _ = kmean_anchors() |
|
|
|
""" |
|
|
|
thr = 1. / thr |
|
|
|
prefix = colorstr('blue', 'bold', 'autoanchor') + ': ' |
|
|
|
|
|
|
|
def metric(k, wh): # compute metrics |
|
|
|
r = wh[:, None] / k[None] |
|
|
@@ -85,9 +89,9 @@ def kmean_anchors(path='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=10 |
|
|
|
k = k[np.argsort(k.prod(1))] # sort small to large |
|
|
|
x, best = metric(k, wh0) |
|
|
|
bpr, aat = (best > thr).float().mean(), (x > thr).float().mean() * n # best possible recall, anch > thr |
|
|
|
print('thr=%.2f: %.4f best possible recall, %.2f anchors past thr' % (thr, bpr, aat)) |
|
|
|
print('n=%g, img_size=%s, metric_all=%.3f/%.3f-mean/best, past_thr=%.3f-mean: ' % |
|
|
|
(n, img_size, x.mean(), best.mean(), x[x > thr].mean()), end='') |
|
|
|
print(f'{prefix}thr={thr:.2f}: {bpr:.4f} best possible recall, {aat:.2f} anchors past thr') |
|
|
|
print(f'{prefix}n={n}, img_size={img_size}, metric_all={x.mean():.3f}/{best.mean():.3f}-mean/best, ' |
|
|
|
f'past_thr={x[x > thr].mean():.3f}-mean: ', end='') |
|
|
|
for i, x in enumerate(k): |
|
|
|
print('%i,%i' % (round(x[0]), round(x[1])), end=', ' if i < len(k) - 1 else '\n') # use in *.cfg |
|
|
|
return k |
|
|
@@ -107,13 +111,12 @@ def kmean_anchors(path='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=10 |
|
|
|
# Filter |
|
|
|
i = (wh0 < 3.0).any(1).sum() |
|
|
|
if i: |
|
|
|
print('WARNING: Extremely small objects found. ' |
|
|
|
'%g of %g labels are < 3 pixels in width or height.' % (i, len(wh0))) |
|
|
|
print(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 |
|
|
|
|
|
|
|
# Kmeans calculation |
|
|
|
print('Running kmeans for %g anchors on %g points...' % (n, len(wh))) |
|
|
|
print(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 |
|
|
|
k *= s |
|
|
@@ -136,7 +139,7 @@ def kmean_anchors(path='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=10 |
|
|
|
# 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='Evolving anchors with Genetic Algorithm') # progress bar |
|
|
|
pbar = tqdm(range(gen), desc=f'{prefix}Evolving anchors with Genetic Algorithm:') # progress bar |
|
|
|
for _ in pbar: |
|
|
|
v = np.ones(sh) |
|
|
|
while (v == 1).all(): # mutate until a change occurs (prevent duplicates) |
|
|
@@ -145,7 +148,7 @@ def kmean_anchors(path='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=10 |
|
|
|
fg = anchor_fitness(kg) |
|
|
|
if fg > f: |
|
|
|
f, k = fg, kg.copy() |
|
|
|
pbar.desc = 'Evolving anchors with Genetic Algorithm: fitness = %.4f' % f |
|
|
|
pbar.desc = f'{prefix}Evolving anchors with Genetic Algorithm: fitness = {f:.4f}' |
|
|
|
if verbose: |
|
|
|
print_results(k) |
|
|
|
|