372 lines
16 KiB
Python
Executable File
372 lines
16 KiB
Python
Executable File
from utils.parse_config import *
|
|
from utils.utils import *
|
|
from pathlib import Path
|
|
|
|
import torch.nn.functional as F
|
|
|
|
ONNX_EXPORT = False
|
|
|
|
|
|
def create_modules(module_defs):
|
|
"""
|
|
Constructs module list of layer blocks from module configuration in module_defs
|
|
"""
|
|
hyperparams = module_defs.pop(0)
|
|
output_filters = [int(hyperparams['channels'])]
|
|
module_list = nn.ModuleList()
|
|
yolo_index = -1
|
|
|
|
for mdef in module_defs:
|
|
modules = nn.Sequential()
|
|
|
|
if mdef['type'] == 'convolutional':
|
|
bn = int(mdef['batch_normalize'])
|
|
filters = int(mdef['filters'])
|
|
kernel_size = int(mdef['size'])
|
|
pad = (kernel_size - 1) // 2 if int(mdef['pad']) else 0
|
|
modules.add_module('Conv2d', nn.Conv2d(in_channels=output_filters[-1],
|
|
out_channels=filters,
|
|
kernel_size=kernel_size,
|
|
stride=int(mdef['stride']),
|
|
padding=pad,
|
|
bias=not bn))
|
|
if bn:
|
|
modules.add_module('BatchNorm2d', nn.BatchNorm2d(filters, momentum=0.1))
|
|
if mdef['activation'] == 'leaky':
|
|
# modules.add_module('activation', nn.PReLU(num_parameters=filters, init=0.1))
|
|
modules.add_module('activation', nn.LeakyReLU(0.1, inplace=True))
|
|
|
|
elif mdef['type'] == 'maxpool':
|
|
kernel_size = int(mdef['size'])
|
|
stride = int(mdef['stride'])
|
|
maxpool = nn.MaxPool2d(kernel_size=kernel_size, stride=stride, padding=int((kernel_size - 1) // 2))
|
|
if kernel_size == 2 and stride == 1: # yolov3-tiny
|
|
modules.add_module('ZeroPad2d', nn.ZeroPad2d((0, 1, 0, 1)))
|
|
modules.add_module('MaxPool2d', maxpool)
|
|
else:
|
|
modules = maxpool
|
|
|
|
elif mdef['type'] == 'upsample':
|
|
modules = nn.Upsample(scale_factor=int(mdef['stride']), mode='nearest')
|
|
|
|
elif mdef['type'] == 'route': # nn.Sequential() placeholder for 'route' layer
|
|
layers = [int(x) for x in mdef['layers'].split(',')]
|
|
filters = sum([output_filters[i + 1 if i > 0 else i] for i in layers])
|
|
# if mdef[i+1]['type'] == 'reorg3d':
|
|
# modules = nn.Upsample(scale_factor=1/float(mdef[i+1]['stride']), mode='nearest') # reorg3d
|
|
|
|
elif mdef['type'] == 'shortcut': # nn.Sequential() placeholder for 'shortcut' layer
|
|
filters = output_filters[int(mdef['from'])]
|
|
|
|
elif mdef['type'] == 'reorg3d': # yolov3-spp-pan-scale
|
|
# torch.Size([16, 128, 104, 104])
|
|
# torch.Size([16, 64, 208, 208]) <-- # stride 2 interpolate dimensions 2 and 3 to cat with prior layer
|
|
pass
|
|
|
|
elif mdef['type'] == 'yolo':
|
|
yolo_index += 1
|
|
mask = [int(x) for x in mdef['mask'].split(',')] # anchor mask
|
|
a = [float(x) for x in mdef['anchors'].split(',')] # anchors
|
|
a = [(a[i], a[i + 1]) for i in range(0, len(a), 2)]
|
|
modules = YOLOLayer(anchors=[a[i] for i in mask], # anchor list
|
|
nc=int(mdef['classes']), # number of classes
|
|
img_size=hyperparams['height'], # 416
|
|
yolo_index=yolo_index) # 0, 1 or 2
|
|
else:
|
|
print('Warning: Unrecognized Layer Type: ' + mdef['type'])
|
|
|
|
# Register module list and number of output filters
|
|
module_list.append(modules)
|
|
output_filters.append(filters)
|
|
|
|
return hyperparams, module_list
|
|
|
|
|
|
class YOLOLayer(nn.Module):
|
|
def __init__(self, anchors, nc, img_size, yolo_index):
|
|
super(YOLOLayer, self).__init__()
|
|
|
|
self.anchors = torch.Tensor(anchors)
|
|
self.na = len(anchors) # number of anchors (3)
|
|
self.nc = nc # number of classes (80)
|
|
self.nx = 0 # initialize number of x gridpoints
|
|
self.ny = 0 # initialize number of y gridpoints
|
|
|
|
if ONNX_EXPORT: # grids must be computed in __init__
|
|
stride = [32, 16, 8][yolo_index] # stride of this layer
|
|
nx = int(img_size[1] / stride) # number x grid points
|
|
ny = int(img_size[0] / stride) # number y grid points
|
|
create_grids(self, max(img_size), (nx, ny))
|
|
|
|
def forward(self, p, img_size, var=None):
|
|
if ONNX_EXPORT:
|
|
bs = 1 # batch size
|
|
else:
|
|
bs, ny, nx = p.shape[0], p.shape[-2], p.shape[-1]
|
|
if (self.nx, self.ny) != (nx, ny):
|
|
create_grids(self, img_size, (nx, ny), p.device, p.dtype)
|
|
|
|
# p.view(bs, 255, 13, 13) -- > (bs, 3, 13, 13, 85) # (bs, anchors, grid, grid, classes + xywh)
|
|
p = p.view(bs, self.na, self.nc + 5, self.ny, self.nx).permute(0, 1, 3, 4, 2).contiguous() # prediction
|
|
|
|
if self.training:
|
|
return p
|
|
|
|
elif ONNX_EXPORT:
|
|
# Constants CAN NOT BE BROADCAST, ensure correct shape!
|
|
ngu = self.ng.repeat((1, self.na * self.nx * self.ny, 1))
|
|
grid_xy = self.grid_xy.repeat((1, self.na, 1, 1, 1)).view((1, -1, 2))
|
|
anchor_wh = self.anchor_wh.repeat((1, 1, self.nx, self.ny, 1)).view((1, -1, 2)) / ngu
|
|
|
|
# p = p.view(-1, 5 + self.nc)
|
|
# xy = torch.sigmoid(p[..., 0:2]) + grid_xy[0] # x, y
|
|
# wh = torch.exp(p[..., 2:4]) * anchor_wh[0] # width, height
|
|
# p_conf = torch.sigmoid(p[:, 4:5]) # Conf
|
|
# p_cls = F.softmax(p[:, 5:85], 1) * p_conf # SSD-like conf
|
|
# return torch.cat((xy / ngu[0], wh, p_conf, p_cls), 1).t()
|
|
|
|
p = p.view(1, -1, 5 + self.nc)
|
|
xy = torch.sigmoid(p[..., 0:2]) + grid_xy # x, y
|
|
wh = torch.exp(p[..., 2:4]) * anchor_wh # width, height
|
|
p_conf = torch.sigmoid(p[..., 4:5]) # Conf
|
|
p_cls = p[..., 5:5 + self.nc]
|
|
# Broadcasting only supported on first dimension in CoreML. See onnx-coreml/_operators.py
|
|
# p_cls = F.softmax(p_cls, 2) * p_conf # SSD-like conf
|
|
p_cls = torch.exp(p_cls).permute((2, 1, 0))
|
|
p_cls = p_cls / p_cls.sum(0).unsqueeze(0) * p_conf.permute((2, 1, 0)) # F.softmax() equivalent
|
|
p_cls = p_cls.permute(2, 1, 0)
|
|
return torch.cat((xy / ngu, wh, p_conf, p_cls), 2).squeeze().t()
|
|
|
|
else: # inference
|
|
# s = 1.5 # scale_xy (pxy = pxy * s - (s - 1) / 2)
|
|
io = p.clone() # inference output
|
|
io[..., 0:2] = torch.sigmoid(io[..., 0:2]) + self.grid_xy # xy
|
|
io[..., 2:4] = torch.exp(io[..., 2:4]) * self.anchor_wh # wh yolo method
|
|
io[..., :4] *= self.stride
|
|
|
|
# io[..., 2:4] = ((torch.sigmoid(io[..., 2:4]) * 2) ** 3) * self.anchor_wh # wh power method
|
|
io[..., 4:] = torch.sigmoid(io[..., 4:]) # p_conf, p_cls
|
|
# io[..., 5:] = F.softmax(io[..., 5:], dim=4) # p_cls
|
|
|
|
if self.nc == 1:
|
|
io[..., 5] = 1 # single-class model https://github.com/ultralytics/yolov3/issues/235
|
|
|
|
# reshape from [1, 3, 13, 13, 85] to [1, 507, 85]
|
|
return io.view(bs, -1, 5 + self.nc), p
|
|
|
|
|
|
class Darknet(nn.Module):
|
|
"""YOLOv3 object detection model"""
|
|
|
|
def __init__(self, cfg, img_size=(416, 416)):
|
|
super(Darknet, self).__init__()
|
|
|
|
self.module_defs = parse_model_cfg(cfg)
|
|
self.module_defs[0]['cfg'] = cfg
|
|
self.module_defs[0]['height'] = img_size
|
|
self.hyperparams, self.module_list = create_modules(self.module_defs)
|
|
self.yolo_layers = get_yolo_layers(self)
|
|
|
|
# Darknet Header https://github.com/AlexeyAB/darknet/issues/2914#issuecomment-496675346
|
|
self.version = np.array([0, 2, 5], dtype=np.int32) # (int32) version info: major, minor, revision
|
|
self.seen = np.array([0], dtype=np.int64) # (int64) number of images seen during training
|
|
|
|
def forward(self, x, var=None):
|
|
img_size = max(x.shape[-2:])
|
|
layer_outputs = []
|
|
output = []
|
|
|
|
for i, (mdef, module) in enumerate(zip(self.module_defs, self.module_list)):
|
|
mtype = mdef['type']
|
|
if mtype in ['convolutional', 'upsample', 'maxpool']:
|
|
x = module(x)
|
|
elif mtype == 'route':
|
|
layer_i = [int(x) for x in mdef['layers'].split(',')]
|
|
if len(layer_i) == 1:
|
|
x = layer_outputs[layer_i[0]]
|
|
else:
|
|
try:
|
|
x = torch.cat([layer_outputs[i] for i in layer_i], 1)
|
|
except: # apply stride 2 for darknet reorg layer
|
|
layer_outputs[layer_i[1]] = F.interpolate(layer_outputs[layer_i[1]], scale_factor=[0.5, 0.5])
|
|
x = torch.cat([layer_outputs[i] for i in layer_i], 1)
|
|
# print(''), [print(layer_outputs[i].shape) for i in layer_i], print(x.shape)
|
|
elif mtype == 'shortcut':
|
|
layer_i = int(mdef['from'])
|
|
x = layer_outputs[-1] + layer_outputs[layer_i]
|
|
elif mtype == 'yolo':
|
|
x = module(x, img_size)
|
|
output.append(x)
|
|
layer_outputs.append(x)
|
|
|
|
if self.training:
|
|
return output
|
|
elif ONNX_EXPORT:
|
|
output = torch.cat(output, 1) # cat 3 layers 85 x (507, 2028, 8112) to 85 x 10647
|
|
nc = self.module_list[self.yolo_layers[0]].nc # number of classes
|
|
return output[5:5 + nc].t(), output[:4].t() # ONNX scores, boxes
|
|
else:
|
|
io, p = list(zip(*output)) # inference output, training output
|
|
return torch.cat(io, 1), p
|
|
|
|
def fuse(self):
|
|
# Fuse Conv2d + BatchNorm2d layers throughout model
|
|
fused_list = nn.ModuleList()
|
|
for a in list(self.children())[0]:
|
|
if isinstance(a, nn.Sequential):
|
|
for i, b in enumerate(a):
|
|
if isinstance(b, nn.modules.batchnorm.BatchNorm2d):
|
|
# fuse this bn layer with the previous conv2d layer
|
|
conv = a[i - 1]
|
|
fused = torch_utils.fuse_conv_and_bn(conv, b)
|
|
a = nn.Sequential(fused, *list(a.children())[i + 1:])
|
|
break
|
|
fused_list.append(a)
|
|
self.module_list = fused_list
|
|
# model_info(self) # yolov3-spp reduced from 225 to 152 layers
|
|
|
|
|
|
def get_yolo_layers(model):
|
|
return [i for i, x in enumerate(model.module_defs) if x['type'] == 'yolo'] # [82, 94, 106] for yolov3
|
|
|
|
|
|
def create_grids(self, img_size=416, ng=(13, 13), device='cpu', type=torch.float32):
|
|
nx, ny = ng # x and y grid size
|
|
self.img_size = img_size
|
|
self.stride = img_size / max(ng)
|
|
|
|
# build xy offsets
|
|
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
|
|
self.grid_xy = torch.stack((xv, yv), 2).to(device).type(type).view((1, 1, ny, nx, 2))
|
|
|
|
# build wh gains
|
|
self.anchor_vec = self.anchors.to(device) / self.stride
|
|
self.anchor_wh = self.anchor_vec.view(1, self.na, 1, 1, 2).to(device).type(type)
|
|
self.ng = torch.Tensor(ng).to(device)
|
|
self.nx = nx
|
|
self.ny = ny
|
|
|
|
|
|
def load_darknet_weights(self, weights, cutoff=-1):
|
|
# Parses and loads the weights stored in 'weights'
|
|
# cutoff: save layers between 0 and cutoff (if cutoff = -1 all are saved)
|
|
file = Path(weights).name
|
|
|
|
# Try to download weights if not available locally
|
|
msg = weights + ' missing, download from https://drive.google.com/drive/folders/1uxgUBemJVw9wZsdpboYbzUN4bcRhsuAI'
|
|
if not os.path.isfile(weights):
|
|
try:
|
|
url = 'https://pjreddie.com/media/files/' + file
|
|
print('Downloading ' + url)
|
|
os.system('curl -f ' + url + ' -o ' + weights)
|
|
except IOError:
|
|
print(msg)
|
|
assert os.path.exists(weights), msg # download missing weights from Google Drive
|
|
|
|
# Establish cutoffs
|
|
if file == 'darknet53.conv.74':
|
|
cutoff = 75
|
|
elif file == 'yolov3-tiny.conv.15':
|
|
cutoff = 15
|
|
|
|
# Read weights file
|
|
with open(weights, 'rb') as f:
|
|
# Read Header https://github.com/AlexeyAB/darknet/issues/2914#issuecomment-496675346
|
|
self.version = np.fromfile(f, dtype=np.int32, count=3) # (int32) version info: major, minor, revision
|
|
self.seen = np.fromfile(f, dtype=np.int64, count=1) # (int64) number of images seen during training
|
|
|
|
weights = np.fromfile(f, dtype=np.float32) # The rest are weights
|
|
|
|
ptr = 0
|
|
for i, (mdef, module) in enumerate(zip(self.module_defs[:cutoff], self.module_list[:cutoff])):
|
|
if mdef['type'] == 'convolutional':
|
|
conv_layer = module[0]
|
|
if mdef['batch_normalize']:
|
|
# Load BN bias, weights, running mean and running variance
|
|
bn_layer = module[1]
|
|
num_b = bn_layer.bias.numel() # Number of biases
|
|
# Bias
|
|
bn_b = torch.from_numpy(weights[ptr:ptr + num_b]).view_as(bn_layer.bias)
|
|
bn_layer.bias.data.copy_(bn_b)
|
|
ptr += num_b
|
|
# Weight
|
|
bn_w = torch.from_numpy(weights[ptr:ptr + num_b]).view_as(bn_layer.weight)
|
|
bn_layer.weight.data.copy_(bn_w)
|
|
ptr += num_b
|
|
# Running Mean
|
|
bn_rm = torch.from_numpy(weights[ptr:ptr + num_b]).view_as(bn_layer.running_mean)
|
|
bn_layer.running_mean.data.copy_(bn_rm)
|
|
ptr += num_b
|
|
# Running Var
|
|
bn_rv = torch.from_numpy(weights[ptr:ptr + num_b]).view_as(bn_layer.running_var)
|
|
bn_layer.running_var.data.copy_(bn_rv)
|
|
ptr += num_b
|
|
else:
|
|
# Load conv. bias
|
|
num_b = conv_layer.bias.numel()
|
|
conv_b = torch.from_numpy(weights[ptr:ptr + num_b]).view_as(conv_layer.bias)
|
|
conv_layer.bias.data.copy_(conv_b)
|
|
ptr += num_b
|
|
# Load conv. weights
|
|
num_w = conv_layer.weight.numel()
|
|
conv_w = torch.from_numpy(weights[ptr:ptr + num_w]).view_as(conv_layer.weight)
|
|
conv_layer.weight.data.copy_(conv_w)
|
|
ptr += num_w
|
|
|
|
return cutoff
|
|
|
|
|
|
def save_weights(self, path='model.weights', cutoff=-1):
|
|
# Converts a PyTorch model to Darket format (*.pt to *.weights)
|
|
# Note: Does not work if model.fuse() is applied
|
|
with open(path, 'wb') as f:
|
|
# Write Header https://github.com/AlexeyAB/darknet/issues/2914#issuecomment-496675346
|
|
self.version.tofile(f) # (int32) version info: major, minor, revision
|
|
self.seen.tofile(f) # (int64) number of images seen during training
|
|
|
|
# Iterate through layers
|
|
for i, (mdef, module) in enumerate(zip(self.module_defs[:cutoff], self.module_list[:cutoff])):
|
|
if mdef['type'] == 'convolutional':
|
|
conv_layer = module[0]
|
|
# If batch norm, load bn first
|
|
if mdef['batch_normalize']:
|
|
bn_layer = module[1]
|
|
bn_layer.bias.data.cpu().numpy().tofile(f)
|
|
bn_layer.weight.data.cpu().numpy().tofile(f)
|
|
bn_layer.running_mean.data.cpu().numpy().tofile(f)
|
|
bn_layer.running_var.data.cpu().numpy().tofile(f)
|
|
# Load conv bias
|
|
else:
|
|
conv_layer.bias.data.cpu().numpy().tofile(f)
|
|
# Load conv weights
|
|
conv_layer.weight.data.cpu().numpy().tofile(f)
|
|
|
|
|
|
def convert(cfg='cfg/yolov3-spp.cfg', weights='weights/yolov3-spp.weights'):
|
|
# Converts between PyTorch and Darknet format per extension (i.e. *.weights convert to *.pt and vice versa)
|
|
# from models import *; convert('cfg/yolov3-spp.cfg', 'weights/yolov3-spp.weights')
|
|
|
|
# Initialize model
|
|
model = Darknet(cfg)
|
|
|
|
# Load weights and save
|
|
if weights.endswith('.pt'): # if PyTorch format
|
|
model.load_state_dict(torch.load(weights, map_location='cpu')['model'])
|
|
save_weights(model, path='converted.weights', cutoff=-1)
|
|
print("Success: converted '%s' to 'converted.weights'" % weights)
|
|
|
|
elif weights.endswith('.weights'): # darknet format
|
|
_ = load_darknet_weights(model, weights)
|
|
|
|
chkpt = {'epoch': -1,
|
|
'best_fitness': None,
|
|
'training_results': None,
|
|
'model': model.state_dict(),
|
|
'optimizer': None}
|
|
|
|
torch.save(chkpt, 'converted.pt')
|
|
print("Success: converted '%s' to 'converted.pt'" % weights)
|
|
|
|
else:
|
|
print('Error: extension not supported.')
|