apoorva/greenhouse
2
1
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

YOLOv5 v6.0 compatibility update (#1857)

Browse Source 
* Initial commit

* Initial commit

* Cleanup

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Fix precommit errors

* Remove TF builds from CI

* export last.pt

* Created using Colaboratory

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
Glenn Jocher
2021-11-14 22:22:59 +01:00
committed by GitHub
parent 1be31704c9
commit 7eb23e3c1d
90 changed files with 6642 additions and 4145 deletions
Download Patch File Download Diff File Expand all files Collapse all files
models/common.py
+285 -89
View File
@@ -1,9 +1,16 @@
# YOLOv3 common modules
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
"""
Common modules
"""
import json
import math
import platform
import warnings
from copy import copy
from pathlib import Path
import cv2
import numpy as np
import pandas as pd
import requests
@@ -12,10 +19,11 @@ import torch.nn as nn
from PIL import Image
from torch.cuda import amp
from utils.datasets import letterbox
from utils.general import non_max_suppression, make_divisible, scale_coords, increment_path, xyxy2xywh, save_one_box
from utils.plots import colors, plot_one_box
from utils.torch_utils import time_synchronized
from utils.datasets import exif_transpose, letterbox
from utils.general import (LOGGER, check_requirements, check_suffix, colorstr, increment_path, make_divisible,
non_max_suppression, scale_coords, xywh2xyxy, xyxy2xywh)
from utils.plots import Annotator, colors, save_one_box
from utils.torch_utils import time_sync
def autopad(k, p=None): # kernel, padding
@@ -25,26 +33,27 @@ def autopad(k, p=None): # kernel, padding
return p
def DWConv(c1, c2, k=1, s=1, act=True):
# Depthwise convolution
return Conv(c1, c2, k, s, g=math.gcd(c1, c2), act=act)
class Conv(nn.Module):
# Standard convolution
def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups
super(Conv, self).__init__()
super().__init__()
self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g, bias=False)
self.bn = nn.BatchNorm2d(c2)
self.act = nn.LeakyReLU(0.1) if act is True else (act if isinstance(act, nn.Module) else nn.Identity())
self.act = nn.SiLU() if act is True else (act if isinstance(act, nn.Module) else nn.Identity())
def forward(self, x):
return self.act(self.bn(self.conv(x)))
def fuseforward(self, x):
def forward_fuse(self, x):
return self.act(self.conv(x))
class DWConv(Conv):
# Depth-wise convolution class
def __init__(self, c1, c2, k=1, s=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups
super().__init__(c1, c2, k, s, g=math.gcd(c1, c2), act=act)
class TransformerLayer(nn.Module):
# Transformer layer https://arxiv.org/abs/2010.11929 (LayerNorm layers removed for better performance)
def __init__(self, c, num_heads):
@@ -70,31 +79,21 @@ class TransformerBlock(nn.Module):
if c1 != c2:
self.conv = Conv(c1, c2)
self.linear = nn.Linear(c2, c2) # learnable position embedding
self.tr = nn.Sequential(*[TransformerLayer(c2, num_heads) for _ in range(num_layers)])
self.tr = nn.Sequential(*(TransformerLayer(c2, num_heads) for _ in range(num_layers)))
self.c2 = c2
def forward(self, x):
if self.conv is not None:
x = self.conv(x)
b, _, w, h = x.shape
p = x.flatten(2)
p = p.unsqueeze(0)
p = p.transpose(0, 3)
p = p.squeeze(3)
e = self.linear(p)
x = p + e
x = self.tr(x)
x = x.unsqueeze(3)
x = x.transpose(0, 3)
x = x.reshape(b, self.c2, w, h)
return x
p = x.flatten(2).unsqueeze(0).transpose(0, 3).squeeze(3)
return self.tr(p + self.linear(p)).unsqueeze(3).transpose(0, 3).reshape(b, self.c2, w, h)
class Bottleneck(nn.Module):
# Standard bottleneck
def __init__(self, c1, c2, shortcut=True, g=1, e=0.5): # ch_in, ch_out, shortcut, groups, expansion
super(Bottleneck, self).__init__()
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = Conv(c1, c_, 1, 1)
self.cv2 = Conv(c_, c2, 3, 1, g=g)
@@ -107,15 +106,15 @@ class Bottleneck(nn.Module):
class BottleneckCSP(nn.Module):
# CSP Bottleneck https://github.com/WongKinYiu/CrossStagePartialNetworks
def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, number, shortcut, groups, expansion
super(BottleneckCSP, self).__init__()
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = Conv(c1, c_, 1, 1)
self.cv2 = nn.Conv2d(c1, c_, 1, 1, bias=False)
self.cv3 = nn.Conv2d(c_, c_, 1, 1, bias=False)
self.cv4 = Conv(2 * c_, c2, 1, 1)
self.bn = nn.BatchNorm2d(2 * c_) # applied to cat(cv2, cv3)
self.act = nn.LeakyReLU(0.1, inplace=True)
self.m = nn.Sequential(*[Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n)])
self.act = nn.SiLU()
self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n)))
def forward(self, x):
y1 = self.cv3(self.m(self.cv1(x)))
@@ -126,12 +125,12 @@ class BottleneckCSP(nn.Module):
class C3(nn.Module):
# CSP Bottleneck with 3 convolutions
def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, number, shortcut, groups, expansion
super(C3, self).__init__()
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = Conv(c1, c_, 1, 1)
self.cv2 = Conv(c1, c_, 1, 1)
self.cv3 = Conv(2 * c_, c2, 1) # act=FReLU(c2)
self.m = nn.Sequential(*[Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n)])
self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n)))
# self.m = nn.Sequential(*[CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n)])
def forward(self, x):
@@ -146,10 +145,26 @@ class C3TR(C3):
self.m = TransformerBlock(c_, c_, 4, n)
class C3SPP(C3):
# C3 module with SPP()
def __init__(self, c1, c2, k=(5, 9, 13), n=1, shortcut=True, g=1, e=0.5):
super().__init__(c1, c2, n, shortcut, g, e)
c_ = int(c2 * e)
self.m = SPP(c_, c_, k)
class C3Ghost(C3):
# C3 module with GhostBottleneck()
def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5):
super().__init__(c1, c2, n, shortcut, g, e)
c_ = int(c2 * e) # hidden channels
self.m = nn.Sequential(*(GhostBottleneck(c_, c_) for _ in range(n)))
class SPP(nn.Module):
# Spatial pyramid pooling layer used in YOLOv3-SPP
# Spatial Pyramid Pooling (SPP) layer https://arxiv.org/abs/1406.4729
def __init__(self, c1, c2, k=(5, 9, 13)):
super(SPP, self).__init__()
super().__init__()
c_ = c1 // 2 # hidden channels
self.cv1 = Conv(c1, c_, 1, 1)
self.cv2 = Conv(c_ * (len(k) + 1), c2, 1, 1)
@@ -157,13 +172,33 @@ class SPP(nn.Module):
def forward(self, x):
x = self.cv1(x)
return self.cv2(torch.cat([x] + [m(x) for m in self.m], 1))
with warnings.catch_warnings():
warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning
return self.cv2(torch.cat([x] + [m(x) for m in self.m], 1))
class SPPF(nn.Module):
# Spatial Pyramid Pooling - Fast (SPPF) layer for by Glenn Jocher
def __init__(self, c1, c2, k=5): # equivalent to SPP(k=(5, 9, 13))
super().__init__()
c_ = c1 // 2 # hidden channels
self.cv1 = Conv(c1, c_, 1, 1)
self.cv2 = Conv(c_ * 4, c2, 1, 1)
self.m = nn.MaxPool2d(kernel_size=k, stride=1, padding=k // 2)
def forward(self, x):
x = self.cv1(x)
with warnings.catch_warnings():
warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning
y1 = self.m(x)
y2 = self.m(y1)
return self.cv2(torch.cat([x, y1, y2, self.m(y2)], 1))
class Focus(nn.Module):
# Focus wh information into c-space
def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups
super(Focus, self).__init__()
super().__init__()
self.conv = Conv(c1 * 4, c2, k, s, p, g, act)
# self.contract = Contract(gain=2)
@@ -172,6 +207,34 @@ class Focus(nn.Module):
# return self.conv(self.contract(x))
class GhostConv(nn.Module):
# Ghost Convolution https://github.com/huawei-noah/ghostnet
def __init__(self, c1, c2, k=1, s=1, g=1, act=True): # ch_in, ch_out, kernel, stride, groups
super().__init__()
c_ = c2 // 2 # hidden channels
self.cv1 = Conv(c1, c_, k, s, None, g, act)
self.cv2 = Conv(c_, c_, 5, 1, None, c_, act)
def forward(self, x):
y = self.cv1(x)
return torch.cat([y, self.cv2(y)], 1)
class GhostBottleneck(nn.Module):
# Ghost Bottleneck https://github.com/huawei-noah/ghostnet
def __init__(self, c1, c2, k=3, s=1): # ch_in, ch_out, kernel, stride
super().__init__()
c_ = c2 // 2
self.conv = nn.Sequential(GhostConv(c1, c_, 1, 1), # pw
DWConv(c_, c_, k, s, act=False) if s == 2 else nn.Identity(), # dw
GhostConv(c_, c2, 1, 1, act=False)) # pw-linear
self.shortcut = nn.Sequential(DWConv(c1, c1, k, s, act=False),
Conv(c1, c2, 1, 1, act=False)) if s == 2 else nn.Identity()
def forward(self, x):
return self.conv(x) + self.shortcut(x)
class Contract(nn.Module):
# Contract width-height into channels, i.e. x(1,64,80,80) to x(1,256,40,40)
def __init__(self, gain=2):
@@ -179,11 +242,11 @@ class Contract(nn.Module):
self.gain = gain
def forward(self, x):
N, C, H, W = x.size() # assert (H / s == 0) and (W / s == 0), 'Indivisible gain'
b, c, h, w = x.size() # assert (h / s == 0) and (W / s == 0), 'Indivisible gain'
s = self.gain
x = x.view(N, C, H // s, s, W // s, s) # x(1,64,40,2,40,2)
x = x.view(b, c, h // s, s, w // s, s) # x(1,64,40,2,40,2)
x = x.permute(0, 3, 5, 1, 2, 4).contiguous() # x(1,2,2,64,40,40)
return x.view(N, C * s * s, H // s, W // s) # x(1,256,40,40)
return x.view(b, c * s * s, h // s, w // s) # x(1,256,40,40)
class Expand(nn.Module):
@@ -193,64 +256,183 @@ class Expand(nn.Module):
self.gain = gain
def forward(self, x):
N, C, H, W = x.size() # assert C / s ** 2 == 0, 'Indivisible gain'
b, c, h, w = x.size() # assert C / s ** 2 == 0, 'Indivisible gain'
s = self.gain
x = x.view(N, s, s, C // s ** 2, H, W) # x(1,2,2,16,80,80)
x = x.view(b, s, s, c // s ** 2, h, w) # x(1,2,2,16,80,80)
x = x.permute(0, 3, 4, 1, 5, 2).contiguous() # x(1,16,80,2,80,2)
return x.view(N, C // s ** 2, H * s, W * s) # x(1,16,160,160)
return x.view(b, c // s ** 2, h * s, w * s) # x(1,16,160,160)
class Concat(nn.Module):
# Concatenate a list of tensors along dimension
def __init__(self, dimension=1):
super(Concat, self).__init__()
super().__init__()
self.d = dimension
def forward(self, x):
return torch.cat(x, self.d)
class NMS(nn.Module):
# Non-Maximum Suppression (NMS) module
conf = 0.25 # confidence threshold
iou = 0.45 # IoU threshold
classes = None # (optional list) filter by class
max_det = 1000 # maximum number of detections per image
class DetectMultiBackend(nn.Module):
# MultiBackend class for python inference on various backends
def __init__(self, weights='yolov3.pt', device=None, dnn=True):
# Usage:
# PyTorch: weights = *.pt
# TorchScript: *.torchscript.pt
# CoreML: *.mlmodel
# TensorFlow: *_saved_model
# TensorFlow: *.pb
# TensorFlow Lite: *.tflite
# ONNX Runtime: *.onnx
# OpenCV DNN: *.onnx with dnn=True
super().__init__()
w = str(weights[0] if isinstance(weights, list) else weights)
suffix, suffixes = Path(w).suffix.lower(), ['.pt', '.onnx', '.tflite', '.pb', '', '.mlmodel']
check_suffix(w, suffixes) # check weights have acceptable suffix
pt, onnx, tflite, pb, saved_model, coreml = (suffix == x for x in suffixes) # backend booleans
jit = pt and 'torchscript' in w.lower()
stride, names = 64, [f'class{i}' for i in range(1000)] # assign defaults
def __init__(self):
super(NMS, self).__init__()
if jit: # TorchScript
LOGGER.info(f'Loading {w} for TorchScript inference...')
extra_files = {'config.txt': ''} # model metadata
model = torch.jit.load(w, _extra_files=extra_files)
if extra_files['config.txt']:
d = json.loads(extra_files['config.txt']) # extra_files dict
stride, names = int(d['stride']), d['names']
elif pt: # PyTorch
from models.experimental import attempt_load # scoped to avoid circular import
model = torch.jit.load(w) if 'torchscript' in w else attempt_load(weights, map_location=device)
stride = int(model.stride.max()) # model stride
names = model.module.names if hasattr(model, 'module') else model.names # get class names
elif coreml: # CoreML *.mlmodel
import coremltools as ct
model = ct.models.MLModel(w)
elif dnn: # ONNX OpenCV DNN
LOGGER.info(f'Loading {w} for ONNX OpenCV DNN inference...')
check_requirements(('opencv-python>=4.5.4',))
net = cv2.dnn.readNetFromONNX(w)
elif onnx: # ONNX Runtime
LOGGER.info(f'Loading {w} for ONNX Runtime inference...')
check_requirements(('onnx', 'onnxruntime-gpu' if torch.has_cuda else 'onnxruntime'))
import onnxruntime
session = onnxruntime.InferenceSession(w, None)
else: # TensorFlow model (TFLite, pb, saved_model)
import tensorflow as tf
if pb: # https://www.tensorflow.org/guide/migrate#a_graphpb_or_graphpbtxt
def wrap_frozen_graph(gd, inputs, outputs):
x = tf.compat.v1.wrap_function(lambda: tf.compat.v1.import_graph_def(gd, name=""), []) # wrapped
return x.prune(tf.nest.map_structure(x.graph.as_graph_element, inputs),
tf.nest.map_structure(x.graph.as_graph_element, outputs))
def forward(self, x):
return non_max_suppression(x[0], self.conf, iou_thres=self.iou, classes=self.classes, max_det=self.max_det)
LOGGER.info(f'Loading {w} for TensorFlow *.pb inference...')
graph_def = tf.Graph().as_graph_def()
graph_def.ParseFromString(open(w, 'rb').read())
frozen_func = wrap_frozen_graph(gd=graph_def, inputs="x:0", outputs="Identity:0")
elif saved_model:
LOGGER.info(f'Loading {w} for TensorFlow saved_model inference...')
model = tf.keras.models.load_model(w)
elif tflite: # https://www.tensorflow.org/lite/guide/python#install_tensorflow_lite_for_python
if 'edgetpu' in w.lower():
LOGGER.info(f'Loading {w} for TensorFlow Edge TPU inference...')
import tflite_runtime.interpreter as tfli
delegate = {'Linux': 'libedgetpu.so.1', # install https://coral.ai/software/#edgetpu-runtime
'Darwin': 'libedgetpu.1.dylib',
'Windows': 'edgetpu.dll'}[platform.system()]
interpreter = tfli.Interpreter(model_path=w, experimental_delegates=[tfli.load_delegate(delegate)])
else:
LOGGER.info(f'Loading {w} for TensorFlow Lite inference...')
interpreter = tf.lite.Interpreter(model_path=w) # load TFLite model
interpreter.allocate_tensors() # allocate
input_details = interpreter.get_input_details() # inputs
output_details = interpreter.get_output_details() # outputs
self.__dict__.update(locals()) # assign all variables to self
def forward(self, im, augment=False, visualize=False, val=False):
# MultiBackend inference
b, ch, h, w = im.shape # batch, channel, height, width
if self.pt: # PyTorch
y = self.model(im) if self.jit else self.model(im, augment=augment, visualize=visualize)
return y if val else y[0]
elif self.coreml: # CoreML *.mlmodel
im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3)
im = Image.fromarray((im[0] * 255).astype('uint8'))
# im = im.resize((192, 320), Image.ANTIALIAS)
y = self.model.predict({'image': im}) # coordinates are xywh normalized
box = xywh2xyxy(y['coordinates'] * [[w, h, w, h]]) # xyxy pixels
conf, cls = y['confidence'].max(1), y['confidence'].argmax(1).astype(np.float)
y = np.concatenate((box, conf.reshape(-1, 1), cls.reshape(-1, 1)), 1)
elif self.onnx: # ONNX
im = im.cpu().numpy() # torch to numpy
if self.dnn: # ONNX OpenCV DNN
self.net.setInput(im)
y = self.net.forward()
else: # ONNX Runtime
y = self.session.run([self.session.get_outputs()[0].name], {self.session.get_inputs()[0].name: im})[0]
else: # TensorFlow model (TFLite, pb, saved_model)
im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3)
if self.pb:
y = self.frozen_func(x=self.tf.constant(im)).numpy()
elif self.saved_model:
y = self.model(im, training=False).numpy()
elif self.tflite:
input, output = self.input_details[0], self.output_details[0]
int8 = input['dtype'] == np.uint8 # is TFLite quantized uint8 model
if int8:
scale, zero_point = input['quantization']
im = (im / scale + zero_point).astype(np.uint8) # de-scale
self.interpreter.set_tensor(input['index'], im)
self.interpreter.invoke()
y = self.interpreter.get_tensor(output['index'])
if int8:
scale, zero_point = output['quantization']
y = (y.astype(np.float32) - zero_point) * scale # re-scale
y[..., 0] *= w # x
y[..., 1] *= h # y
y[..., 2] *= w # w
y[..., 3] *= h # h
y = torch.tensor(y)
return (y, []) if val else y
class AutoShape(nn.Module):
# input-robust model wrapper for passing cv2/np/PIL/torch inputs. Includes preprocessing, inference and NMS
# input-robust model wrapper for passing cv2/np/PIL/torch inputs. Includes preprocessing, inference and NMS
conf = 0.25 # NMS confidence threshold
iou = 0.45 # NMS IoU threshold
classes = None # (optional list) filter by class
classes = None # (optional list) filter by class, i.e. = [0, 15, 16] for COCO persons, cats and dogs
multi_label = False # NMS multiple labels per box
max_det = 1000 # maximum number of detections per image
def __init__(self, model):
super(AutoShape, self).__init__()
super().__init__()
self.model = model.eval()
def autoshape(self):
print('AutoShape already enabled, skipping... ') # model already converted to model.autoshape()
LOGGER.info('AutoShape already enabled, skipping... ') # model already converted to model.autoshape()
return self
def _apply(self, fn):
# Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers
self = super()._apply(fn)
m = self.model.model[-1] # Detect()
m.stride = fn(m.stride)
m.grid = list(map(fn, m.grid))
if isinstance(m.anchor_grid, list):
m.anchor_grid = list(map(fn, m.anchor_grid))
return self
@torch.no_grad()
def forward(self, imgs, size=640, augment=False, profile=False):
# Inference from various sources. For height=640, width=1280, RGB images example inputs are:
# filename: imgs = 'data/images/zidane.jpg'
# URI: = 'https://github.com/ultralytics/yolov5/releases/download/v1.0/zidane.jpg'
# file: imgs = 'data/images/zidane.jpg' # str or PosixPath
# URI: = 'https://ultralytics.com/images/zidane.jpg'
# OpenCV: = cv2.imread('image.jpg')[:,:,::-1] # HWC BGR to RGB x(640,1280,3)
# PIL: = Image.open('image.jpg') # HWC x(640,1280,3)
# PIL: = Image.open('image.jpg') or ImageGrab.grab() # HWC x(640,1280,3)
# numpy: = np.zeros((640,1280,3)) # HWC
# torch: = torch.zeros(16,3,320,640) # BCHW (scaled to size=640, 0-1 values)
# multiple: = [Image.open('image1.jpg'), Image.open('image2.jpg'), ...] # list of images
t = [time_synchronized()]
t = [time_sync()]
p = next(self.model.parameters()) # for device and type
if isinstance(imgs, torch.Tensor): # torch
with amp.autocast(enabled=p.device.type != 'cpu'):
@@ -261,14 +443,15 @@ class AutoShape(nn.Module):
shape0, shape1, files = [], [], [] # image and inference shapes, filenames
for i, im in enumerate(imgs):
f = f'image{i}' # filename
if isinstance(im, str): # filename or uri
im, f = np.asarray(Image.open(requests.get(im, stream=True).raw if im.startswith('http') else im)), im
if isinstance(im, (str, Path)): # filename or uri
im, f = Image.open(requests.get(im, stream=True).raw if str(im).startswith('http') else im), im
im = np.asarray(exif_transpose(im))
elif isinstance(im, Image.Image): # PIL Image
im, f = np.asarray(im), getattr(im, 'filename', f) or f
im, f = np.asarray(exif_transpose(im)), getattr(im, 'filename', f) or f
files.append(Path(f).with_suffix('.jpg').name)
if im.shape[0] < 5: # image in CHW
im = im.transpose((1, 2, 0)) # reverse dataloader .transpose(2, 0, 1)
im = im[:, :, :3] if im.ndim == 3 else np.tile(im[:, :, None], 3) # enforce 3ch input
im = im[..., :3] if im.ndim == 3 else np.tile(im[..., None], 3) # enforce 3ch input
s = im.shape[:2] # HWC
shape0.append(s) # image shape
g = (size / max(s)) # gain
@@ -278,29 +461,30 @@ class AutoShape(nn.Module):
x = [letterbox(im, new_shape=shape1, auto=False)[0] for im in imgs] # pad
x = np.stack(x, 0) if n > 1 else x[0][None] # stack
x = np.ascontiguousarray(x.transpose((0, 3, 1, 2))) # BHWC to BCHW
x = torch.from_numpy(x).to(p.device).type_as(p) / 255. # uint8 to fp16/32
t.append(time_synchronized())
x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32
t.append(time_sync())
with amp.autocast(enabled=p.device.type != 'cpu'):
# Inference
y = self.model(x, augment, profile)[0] # forward
t.append(time_synchronized())
t.append(time_sync())
# Post-process
y = non_max_suppression(y, self.conf, iou_thres=self.iou, classes=self.classes, max_det=self.max_det) # NMS
y = non_max_suppression(y, self.conf, iou_thres=self.iou, classes=self.classes,
multi_label=self.multi_label, max_det=self.max_det) # NMS
for i in range(n):
scale_coords(shape1, y[i][:, :4], shape0[i])
t.append(time_synchronized())
t.append(time_sync())
return Detections(imgs, y, files, t, self.names, x.shape)
class Detections:
# detections class for YOLOv3 inference results
# detections class for inference results
def __init__(self, imgs, pred, files, times=None, names=None, shape=None):
super(Detections, self).__init__()
super().__init__()
d = pred[0].device # device
gn = [torch.tensor([*[im.shape[i] for i in [1, 0, 1, 0]], 1., 1.], device=d) for im in imgs] # normalizations
gn = [torch.tensor([*(im.shape[i] for i in [1, 0, 1, 0]), 1, 1], device=d) for im in imgs] # normalizations
self.imgs = imgs # list of images as numpy arrays
self.pred = pred # list of tensors pred[0] = (xyxy, conf, cls)
self.names = names # class names
@@ -314,47 +498,59 @@ class Detections:
self.s = shape # inference BCHW shape
def display(self, pprint=False, show=False, save=False, crop=False, render=False, save_dir=Path('')):
crops = []
for i, (im, pred) in enumerate(zip(self.imgs, self.pred)):
str = f'image {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} '
if pred is not None:
s = f'image {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} ' # string
if pred.shape[0]:
for c in pred[:, -1].unique():
n = (pred[:, -1] == c).sum() # detections per class
str += f"{n} {self.names[int(c)]}{'s' * (n > 1)}, " # add to string
s += f"{n} {self.names[int(c)]}{'s' * (n > 1)}, " # add to string
if show or save or render or crop:
for *box, conf, cls in pred: # xyxy, confidence, class
annotator = Annotator(im, example=str(self.names))
for *box, conf, cls in reversed(pred): # xyxy, confidence, class
label = f'{self.names[int(cls)]} {conf:.2f}'
if crop:
save_one_box(box, im, file=save_dir / 'crops' / self.names[int(cls)] / self.files[i])
file = save_dir / 'crops' / self.names[int(cls)] / self.files[i] if save else None
crops.append({'box': box, 'conf': conf, 'cls': cls, 'label': label,
'im': save_one_box(box, im, file=file, save=save)})
else: # all others
plot_one_box(box, im, label=label, color=colors(cls))
annotator.box_label(box, label, color=colors(cls))
im = annotator.im
else:
s += '(no detections)'
im = Image.fromarray(im.astype(np.uint8)) if isinstance(im, np.ndarray) else im # from np
if pprint:
print(str.rstrip(', '))
LOGGER.info(s.rstrip(', '))
if show:
im.show(self.files[i]) # show
if save:
f = self.files[i]
im.save(save_dir / f) # save
print(f"{'Saved' * (i == 0)} {f}", end=',' if i < self.n - 1 else f' to {save_dir}\n')
if i == self.n - 1:
LOGGER.info(f"Saved {self.n} image{'s' * (self.n > 1)} to {colorstr('bold', save_dir)}")
if render:
self.imgs[i] = np.asarray(im)
if crop:
if save:
LOGGER.info(f'Saved results to {save_dir}\n')
return crops
def print(self):
self.display(pprint=True) # print results
print(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {tuple(self.s)}' % self.t)
LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {tuple(self.s)}' %
self.t)
def show(self):
self.display(show=True) # show results
def save(self, save_dir='runs/hub/exp'):
save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/hub/exp', mkdir=True) # increment save_dir
def save(self, save_dir='runs/detect/exp'):
save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) # increment save_dir
self.display(save=True, save_dir=save_dir) # save results
def crop(self, save_dir='runs/hub/exp'):
save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/hub/exp', mkdir=True) # increment save_dir
self.display(crop=True, save_dir=save_dir) # crop results
print(f'Saved results to {save_dir}\n')
def crop(self, save=True, save_dir='runs/detect/exp'):
save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) if save else None
return self.display(crop=True, save=save, save_dir=save_dir) # crop results
def render(self):
self.display(render=True) # render results
@@ -385,7 +581,7 @@ class Detections:
class Classify(nn.Module):
# Classification head, i.e. x(b,c1,20,20) to x(b,c2)
def __init__(self, c1, c2, k=1, s=1, p=None, g=1): # ch_in, ch_out, kernel, stride, padding, groups
super(Classify, self).__init__()
super().__init__()
self.aap = nn.AdaptiveAvgPool2d(1) # to x(b,c1,1,1)
self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g) # to x(b,c2,1,1)
self.flat = nn.Flatten()
models/experimental.py
+34 -50
View File
@@ -1,18 +1,22 @@
# YOLOv3 experimental modules
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
"""
Experimental modules
"""
import math
import numpy as np
import torch
import torch.nn as nn
from models.common import Conv, DWConv
from utils.google_utils import attempt_download
from models.common import Conv
from utils.downloads import attempt_download
class CrossConv(nn.Module):
# Cross Convolution Downsample
def __init__(self, c1, c2, k=3, s=1, g=1, e=1.0, shortcut=False):
# ch_in, ch_out, kernel, stride, groups, expansion, shortcut
super(CrossConv, self).__init__()
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = Conv(c1, c_, (1, k), (1, s))
self.cv2 = Conv(c_, c2, (k, 1), (s, 1), g=g)
@@ -25,11 +29,11 @@ class CrossConv(nn.Module):
class Sum(nn.Module):
# Weighted sum of 2 or more layers https://arxiv.org/abs/1911.09070
def __init__(self, n, weight=False): # n: number of inputs
super(Sum, self).__init__()
super().__init__()
self.weight = weight # apply weights boolean
self.iter = range(n - 1) # iter object
if weight:
self.w = nn.Parameter(-torch.arange(1., n) / 2, requires_grad=True) # layer weights
self.w = nn.Parameter(-torch.arange(1.0, n) / 2, requires_grad=True) # layer weights
def forward(self, x):
y = x[0] # no weight
@@ -43,86 +47,66 @@ class Sum(nn.Module):
return y
class GhostConv(nn.Module):
# Ghost Convolution https://github.com/huawei-noah/ghostnet
def __init__(self, c1, c2, k=1, s=1, g=1, act=True): # ch_in, ch_out, kernel, stride, groups
super(GhostConv, self).__init__()
c_ = c2 // 2 # hidden channels
self.cv1 = Conv(c1, c_, k, s, None, g, act)
self.cv2 = Conv(c_, c_, 5, 1, None, c_, act)
def forward(self, x):
y = self.cv1(x)
return torch.cat([y, self.cv2(y)], 1)
class GhostBottleneck(nn.Module):
# Ghost Bottleneck https://github.com/huawei-noah/ghostnet
def __init__(self, c1, c2, k=3, s=1): # ch_in, ch_out, kernel, stride
super(GhostBottleneck, self).__init__()
c_ = c2 // 2
self.conv = nn.Sequential(GhostConv(c1, c_, 1, 1), # pw
DWConv(c_, c_, k, s, act=False) if s == 2 else nn.Identity(), # dw
GhostConv(c_, c2, 1, 1, act=False)) # pw-linear
self.shortcut = nn.Sequential(DWConv(c1, c1, k, s, act=False),
Conv(c1, c2, 1, 1, act=False)) if s == 2 else nn.Identity()
def forward(self, x):
return self.conv(x) + self.shortcut(x)
class MixConv2d(nn.Module):
# Mixed Depthwise Conv https://arxiv.org/abs/1907.09595
def __init__(self, c1, c2, k=(1, 3), s=1, equal_ch=True):
super(MixConv2d, self).__init__()
groups = len(k)
# Mixed Depth-wise Conv https://arxiv.org/abs/1907.09595
def __init__(self, c1, c2, k=(1, 3), s=1, equal_ch=True): # ch_in, ch_out, kernel, stride, ch_strategy
super().__init__()
n = len(k) # number of convolutions
if equal_ch: # equal c_ per group
i = torch.linspace(0, groups - 1E-6, c2).floor() # c2 indices
c_ = [(i == g).sum() for g in range(groups)] # intermediate channels
i = torch.linspace(0, n - 1E-6, c2).floor() # c2 indices
c_ = [(i == g).sum() for g in range(n)] # intermediate channels
else: # equal weight.numel() per group
b = [c2] + [0] * groups
a = np.eye(groups + 1, groups, k=-1)
b = [c2] + [0] * n
a = np.eye(n + 1, n, k=-1)
a -= np.roll(a, 1, axis=1)
a *= np.array(k) ** 2
a[0] = 1
c_ = np.linalg.lstsq(a, b, rcond=None)[0].round() # solve for equal weight indices, ax = b
self.m = nn.ModuleList([nn.Conv2d(c1, int(c_[g]), k[g], s, k[g] // 2, bias=False) for g in range(groups)])
self.m = nn.ModuleList(
[nn.Conv2d(c1, int(c_), k, s, k // 2, groups=math.gcd(c1, int(c_)), bias=False) for k, c_ in zip(k, c_)])
self.bn = nn.BatchNorm2d(c2)
self.act = nn.LeakyReLU(0.1, inplace=True)
self.act = nn.SiLU()
def forward(self, x):
return x + self.act(self.bn(torch.cat([m(x) for m in self.m], 1)))
return self.act(self.bn(torch.cat([m(x) for m in self.m], 1)))
class Ensemble(nn.ModuleList):
# Ensemble of models
def __init__(self):
super(Ensemble, self).__init__()
super().__init__()
def forward(self, x, augment=False):
def forward(self, x, augment=False, profile=False, visualize=False):
y = []
for module in self:
y.append(module(x, augment)[0])
y.append(module(x, augment, profile, visualize)[0])
# y = torch.stack(y).max(0)[0] # max ensemble
# y = torch.stack(y).mean(0) # mean ensemble
y = torch.cat(y, 1) # nms ensemble
return y, None # inference, train output
def attempt_load(weights, map_location=None, inplace=True):
def attempt_load(weights, map_location=None, inplace=True, fuse=True):
from models.yolo import Detect, Model
# Loads an ensemble of models weights=[a,b,c] or a single model weights=[a] or weights=a
model = Ensemble()
for w in weights if isinstance(weights, list) else [weights]:
ckpt = torch.load(attempt_download(w), map_location=map_location) # load
model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().fuse().eval()) # FP32 model
if fuse:
model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().fuse().eval()) # FP32 model
else:
model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().eval()) # without layer fuse
# Compatibility updates
for m in model.modules():
if type(m) in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU, Detect, Model]:
m.inplace = inplace # pytorch 1.7.0 compatibility
if type(m) is Detect:
if not isinstance(m.anchor_grid, list): # new Detect Layer compatibility
delattr(m, 'anchor_grid')
setattr(m, 'anchor_grid', [torch.zeros(1)] * m.nl)
elif type(m) is Conv:
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
models/export.py
-145
View File
@@ -1,145 +0,0 @@
"""Exports a YOLOv3 *.pt model to TorchScript, ONNX, CoreML formats
Usage:
$ python path/to/models/export.py --weights yolov3.pt --img 640 --batch 1
"""
import argparse
import sys
import time
from pathlib import Path
sys.path.append(Path(__file__).parent.parent.absolute().__str__()) # to run '$ python *.py' files in subdirectories
import torch
import torch.nn as nn
from torch.utils.mobile_optimizer import optimize_for_mobile
import models
from models.experimental import attempt_load
from utils.activations import Hardswish, SiLU
from utils.general import colorstr, check_img_size, check_requirements, file_size, set_logging
from utils.torch_utils import select_device
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--weights', type=str, default='./yolov3.pt', help='weights path')
parser.add_argument('--img-size', nargs='+', type=int, default=[640, 640], help='image size') # height, width
parser.add_argument('--batch-size', type=int, default=1, help='batch size')
parser.add_argument('--device', default='cpu', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
parser.add_argument('--include', nargs='+', default=['torchscript', 'onnx', 'coreml'], help='include formats')
parser.add_argument('--half', action='store_true', help='FP16 half-precision export')
parser.add_argument('--inplace', action='store_true', help='set YOLOv3 Detect() inplace=True')
parser.add_argument('--train', action='store_true', help='model.train() mode')
parser.add_argument('--optimize', action='store_true', help='optimize TorchScript for mobile') # TorchScript-only
parser.add_argument('--dynamic', action='store_true', help='dynamic ONNX axes') # ONNX-only
parser.add_argument('--simplify', action='store_true', help='simplify ONNX model') # ONNX-only
parser.add_argument('--opset-version', type=int, default=12, help='ONNX opset version') # ONNX-only
opt = parser.parse_args()
opt.img_size *= 2 if len(opt.img_size) == 1 else 1 # expand
opt.include = [x.lower() for x in opt.include]
print(opt)
set_logging()
t = time.time()
# Load PyTorch model
device = select_device(opt.device)
model = attempt_load(opt.weights, map_location=device) # load FP32 model
labels = model.names
# Checks
gs = int(max(model.stride)) # grid size (max stride)
opt.img_size = [check_img_size(x, gs) for x in opt.img_size] # verify img_size are gs-multiples
assert not (opt.device.lower() == 'cpu' and opt.half), '--half only compatible with GPU export, i.e. use --device 0'
# Input
img = torch.zeros(opt.batch_size, 3, *opt.img_size).to(device) # image size(1,3,320,192) iDetection
# Update model
if opt.half:
img, model = img.half(), model.half() # to FP16
if opt.train:
model.train() # training mode (no grid construction in Detect layer)
for k, m in model.named_modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
if isinstance(m, models.common.Conv): # assign export-friendly activations
if isinstance(m.act, nn.Hardswish):
m.act = Hardswish()
elif isinstance(m.act, nn.SiLU):
m.act = SiLU()
elif isinstance(m, models.yolo.Detect):
m.inplace = opt.inplace
m.onnx_dynamic = opt.dynamic
# m.forward = m.forward_export # assign forward (optional)
for _ in range(2):
y = model(img) # dry runs
print(f"\n{colorstr('PyTorch:')} starting from {opt.weights} ({file_size(opt.weights):.1f} MB)")
# TorchScript export -----------------------------------------------------------------------------------------------
if 'torchscript' in opt.include or 'coreml' in opt.include:
prefix = colorstr('TorchScript:')
try:
print(f'\n{prefix} starting export with torch {torch.__version__}...')
f = opt.weights.replace('.pt', '.torchscript.pt') # filename
ts = torch.jit.trace(model, img, strict=False)
(optimize_for_mobile(ts) if opt.optimize else ts).save(f)
print(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)')
except Exception as e:
print(f'{prefix} export failure: {e}')
# ONNX export ------------------------------------------------------------------------------------------------------
if 'onnx' in opt.include:
prefix = colorstr('ONNX:')
try:
import onnx
print(f'{prefix} starting export with onnx {onnx.__version__}...')
f = opt.weights.replace('.pt', '.onnx') # filename
torch.onnx.export(model, img, f, verbose=False, opset_version=opt.opset_version, input_names=['images'],
training=torch.onnx.TrainingMode.TRAINING if opt.train else torch.onnx.TrainingMode.EVAL,
do_constant_folding=not opt.train,
dynamic_axes={'images': {0: 'batch', 2: 'height', 3: 'width'}, # size(1,3,640,640)
'output': {0: 'batch', 2: 'y', 3: 'x'}} if opt.dynamic else None)
# Checks
model_onnx = onnx.load(f) # load onnx model
onnx.checker.check_model(model_onnx) # check onnx model
# print(onnx.helper.printable_graph(model_onnx.graph)) # print
# Simplify
if opt.simplify:
try:
check_requirements(['onnx-simplifier'])
import onnxsim
print(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...')
model_onnx, check = onnxsim.simplify(
model_onnx,
dynamic_input_shape=opt.dynamic,
input_shapes={'images': list(img.shape)} if opt.dynamic else None)
assert check, 'assert check failed'
onnx.save(model_onnx, f)
except Exception as e:
print(f'{prefix} simplifier failure: {e}')
print(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)')
except Exception as e:
print(f'{prefix} export failure: {e}')
# CoreML export ----------------------------------------------------------------------------------------------------
if 'coreml' in opt.include:
prefix = colorstr('CoreML:')
try:
import coremltools as ct
print(f'{prefix} starting export with coremltools {ct.__version__}...')
assert opt.train, 'CoreML exports should be placed in model.train() mode with `python export.py --train`'
model = ct.convert(ts, inputs=[ct.ImageType('image', shape=img.shape, scale=1 / 255.0, bias=[0, 0, 0])])
f = opt.weights.replace('.pt', '.mlmodel') # filename
model.save(f)
print(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)')
except Exception as e:
print(f'{prefix} export failure: {e}')
# Finish
print(f'\nExport complete ({time.time() - t:.2f}s). Visualize with https://github.com/lutzroeder/netron.')
models/tf.py
+465
View File
@@ -0,0 +1,465 @@
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
"""
TensorFlow, Keras and TFLite versions of
Authored by https://github.com/zldrobit in PR https://github.com/ultralytics/yolov5/pull/1127
Usage:
$ python models/tf.py --weights yolov3.pt
Export:
$ python path/to/export.py --weights yolov3.pt --include saved_model pb tflite tfjs
"""
import argparse
import logging
import sys
from copy import deepcopy
from pathlib import Path
FILE = Path(__file__).resolve()
ROOT = FILE.parents[1] # root directory
if str(ROOT) not in sys.path:
sys.path.append(str(ROOT)) # add ROOT to PATH
# ROOT = ROOT.relative_to(Path.cwd()) # relative
import numpy as np
import tensorflow as tf
import torch
import torch.nn as nn
from tensorflow import keras
from models.common import C3, SPP, SPPF, Bottleneck, BottleneckCSP, Concat, Conv, DWConv, Focus, autopad
from models.experimental import CrossConv, MixConv2d, attempt_load
from models.yolo import Detect
from utils.activations import SiLU
from utils.general import LOGGER, make_divisible, print_args
class TFBN(keras.layers.Layer):
# TensorFlow BatchNormalization wrapper
def __init__(self, w=None):
super().__init__()
self.bn = keras.layers.BatchNormalization(
beta_initializer=keras.initializers.Constant(w.bias.numpy()),
gamma_initializer=keras.initializers.Constant(w.weight.numpy()),
moving_mean_initializer=keras.initializers.Constant(w.running_mean.numpy()),
moving_variance_initializer=keras.initializers.Constant(w.running_var.numpy()),
epsilon=w.eps)
def call(self, inputs):
return self.bn(inputs)
class TFPad(keras.layers.Layer):
def __init__(self, pad):
super().__init__()
self.pad = tf.constant([[0, 0], [pad, pad], [pad, pad], [0, 0]])
def call(self, inputs):
return tf.pad(inputs, self.pad, mode='constant', constant_values=0)
class TFConv(keras.layers.Layer):
# Standard convolution
def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True, w=None):
# ch_in, ch_out, weights, kernel, stride, padding, groups
super().__init__()
assert g == 1, "TF v2.2 Conv2D does not support 'groups' argument"
assert isinstance(k, int), "Convolution with multiple kernels are not allowed."
# TensorFlow convolution padding is inconsistent with PyTorch (e.g. k=3 s=2 'SAME' padding)
# see https://stackoverflow.com/questions/52975843/comparing-conv2d-with-padding-between-tensorflow-and-pytorch
conv = keras.layers.Conv2D(
c2, k, s, 'SAME' if s == 1 else 'VALID', use_bias=False if hasattr(w, 'bn') else True,
kernel_initializer=keras.initializers.Constant(w.conv.weight.permute(2, 3, 1, 0).numpy()),
bias_initializer='zeros' if hasattr(w, 'bn') else keras.initializers.Constant(w.conv.bias.numpy()))
self.conv = conv if s == 1 else keras.Sequential([TFPad(autopad(k, p)), conv])
self.bn = TFBN(w.bn) if hasattr(w, 'bn') else tf.identity
# activations
if isinstance(w.act, nn.LeakyReLU):
self.act = (lambda x: keras.activations.relu(x, alpha=0.1)) if act else tf.identity
elif isinstance(w.act, nn.Hardswish):
self.act = (lambda x: x * tf.nn.relu6(x + 3) * 0.166666667) if act else tf.identity
elif isinstance(w.act, (nn.SiLU, SiLU)):
self.act = (lambda x: keras.activations.swish(x)) if act else tf.identity
else:
raise Exception(f'no matching TensorFlow activation found for {w.act}')
def call(self, inputs):
return self.act(self.bn(self.conv(inputs)))
class TFFocus(keras.layers.Layer):
# Focus wh information into c-space
def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True, w=None):
# ch_in, ch_out, kernel, stride, padding, groups
super().__init__()
self.conv = TFConv(c1 * 4, c2, k, s, p, g, act, w.conv)
def call(self, inputs): # x(b,w,h,c) -> y(b,w/2,h/2,4c)
# inputs = inputs / 255 # normalize 0-255 to 0-1
return self.conv(tf.concat([inputs[:, ::2, ::2, :],
inputs[:, 1::2, ::2, :],
inputs[:, ::2, 1::2, :],
inputs[:, 1::2, 1::2, :]], 3))
class TFBottleneck(keras.layers.Layer):
# Standard bottleneck
def __init__(self, c1, c2, shortcut=True, g=1, e=0.5, w=None): # ch_in, ch_out, shortcut, groups, expansion
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1)
self.cv2 = TFConv(c_, c2, 3, 1, g=g, w=w.cv2)
self.add = shortcut and c1 == c2
def call(self, inputs):
return inputs + self.cv2(self.cv1(inputs)) if self.add else self.cv2(self.cv1(inputs))
class TFConv2d(keras.layers.Layer):
# Substitution for PyTorch nn.Conv2D
def __init__(self, c1, c2, k, s=1, g=1, bias=True, w=None):
super().__init__()
assert g == 1, "TF v2.2 Conv2D does not support 'groups' argument"
self.conv = keras.layers.Conv2D(
c2, k, s, 'VALID', use_bias=bias,
kernel_initializer=keras.initializers.Constant(w.weight.permute(2, 3, 1, 0).numpy()),
bias_initializer=keras.initializers.Constant(w.bias.numpy()) if bias else None, )
def call(self, inputs):
return self.conv(inputs)
class TFBottleneckCSP(keras.layers.Layer):
# CSP Bottleneck https://github.com/WongKinYiu/CrossStagePartialNetworks
def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5, w=None):
# ch_in, ch_out, number, shortcut, groups, expansion
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1)
self.cv2 = TFConv2d(c1, c_, 1, 1, bias=False, w=w.cv2)
self.cv3 = TFConv2d(c_, c_, 1, 1, bias=False, w=w.cv3)
self.cv4 = TFConv(2 * c_, c2, 1, 1, w=w.cv4)
self.bn = TFBN(w.bn)
self.act = lambda x: keras.activations.relu(x, alpha=0.1)
self.m = keras.Sequential([TFBottleneck(c_, c_, shortcut, g, e=1.0, w=w.m[j]) for j in range(n)])
def call(self, inputs):
y1 = self.cv3(self.m(self.cv1(inputs)))
y2 = self.cv2(inputs)
return self.cv4(self.act(self.bn(tf.concat((y1, y2), axis=3))))
class TFC3(keras.layers.Layer):
# CSP Bottleneck with 3 convolutions
def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5, w=None):
# ch_in, ch_out, number, shortcut, groups, expansion
super().__init__()
c_ = int(c2 * e) # hidden channels
self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1)
self.cv2 = TFConv(c1, c_, 1, 1, w=w.cv2)
self.cv3 = TFConv(2 * c_, c2, 1, 1, w=w.cv3)
self.m = keras.Sequential([TFBottleneck(c_, c_, shortcut, g, e=1.0, w=w.m[j]) for j in range(n)])
def call(self, inputs):
return self.cv3(tf.concat((self.m(self.cv1(inputs)), self.cv2(inputs)), axis=3))
class TFSPP(keras.layers.Layer):
# Spatial pyramid pooling layer used in YOLOv3-SPP
def __init__(self, c1, c2, k=(5, 9, 13), w=None):
super().__init__()
c_ = c1 // 2 # hidden channels
self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1)
self.cv2 = TFConv(c_ * (len(k) + 1), c2, 1, 1, w=w.cv2)
self.m = [keras.layers.MaxPool2D(pool_size=x, strides=1, padding='SAME') for x in k]
def call(self, inputs):
x = self.cv1(inputs)
return self.cv2(tf.concat([x] + [m(x) for m in self.m], 3))
class TFSPPF(keras.layers.Layer):
# Spatial pyramid pooling-Fast layer
def __init__(self, c1, c2, k=5, w=None):
super().__init__()
c_ = c1 // 2 # hidden channels
self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1)
self.cv2 = TFConv(c_ * 4, c2, 1, 1, w=w.cv2)
self.m = keras.layers.MaxPool2D(pool_size=k, strides=1, padding='SAME')
def call(self, inputs):
x = self.cv1(inputs)
y1 = self.m(x)
y2 = self.m(y1)
return self.cv2(tf.concat([x, y1, y2, self.m(y2)], 3))
class TFDetect(keras.layers.Layer):
def __init__(self, nc=80, anchors=(), ch=(), imgsz=(640, 640), w=None): # detection layer
super().__init__()
self.stride = tf.convert_to_tensor(w.stride.numpy(), dtype=tf.float32)
self.nc = nc # number of classes
self.no = nc + 5 # number of outputs per anchor
self.nl = len(anchors) # number of detection layers
self.na = len(anchors[0]) // 2 # number of anchors
self.grid = [tf.zeros(1)] * self.nl # init grid
self.anchors = tf.convert_to_tensor(w.anchors.numpy(), dtype=tf.float32)
self.anchor_grid = tf.reshape(self.anchors * tf.reshape(self.stride, [self.nl, 1, 1]),
[self.nl, 1, -1, 1, 2])
self.m = [TFConv2d(x, self.no * self.na, 1, w=w.m[i]) for i, x in enumerate(ch)]
self.training = False # set to False after building model
self.imgsz = imgsz
for i in range(self.nl):
ny, nx = self.imgsz[0] // self.stride[i], self.imgsz[1] // self.stride[i]
self.grid[i] = self._make_grid(nx, ny)
def call(self, inputs):
z = [] # inference output
x = []
for i in range(self.nl):
x.append(self.m[i](inputs[i]))
# x(bs,20,20,255) to x(bs,3,20,20,85)
ny, nx = self.imgsz[0] // self.stride[i], self.imgsz[1] // self.stride[i]
x[i] = tf.transpose(tf.reshape(x[i], [-1, ny * nx, self.na, self.no]), [0, 2, 1, 3])
if not self.training: # inference
y = tf.sigmoid(x[i])
xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy
wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i]
# Normalize xywh to 0-1 to reduce calibration error
xy /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32)
wh /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32)
y = tf.concat([xy, wh, y[..., 4:]], -1)
z.append(tf.reshape(y, [-1, 3 * ny * nx, self.no]))
return x if self.training else (tf.concat(z, 1), x)
@staticmethod
def _make_grid(nx=20, ny=20):
# yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
# return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
xv, yv = tf.meshgrid(tf.range(nx), tf.range(ny))
return tf.cast(tf.reshape(tf.stack([xv, yv], 2), [1, 1, ny * nx, 2]), dtype=tf.float32)
class TFUpsample(keras.layers.Layer):
def __init__(self, size, scale_factor, mode, w=None): # warning: all arguments needed including 'w'
super().__init__()
assert scale_factor == 2, "scale_factor must be 2"
self.upsample = lambda x: tf.image.resize(x, (x.shape[1] * 2, x.shape[2] * 2), method=mode)
# self.upsample = keras.layers.UpSampling2D(size=scale_factor, interpolation=mode)
# with default arguments: align_corners=False, half_pixel_centers=False
# self.upsample = lambda x: tf.raw_ops.ResizeNearestNeighbor(images=x,
# size=(x.shape[1] * 2, x.shape[2] * 2))
def call(self, inputs):
return self.upsample(inputs)
class TFConcat(keras.layers.Layer):
def __init__(self, dimension=1, w=None):
super().__init__()
assert dimension == 1, "convert only NCHW to NHWC concat"
self.d = 3
def call(self, inputs):
return tf.concat(inputs, self.d)
def parse_model(d, ch, model, imgsz): # model_dict, input_channels(3)
LOGGER.info(f"\n{'':>3}{'from':>18}{'n':>3}{'params':>10} {'module':<40}{'arguments':<30}")
anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple']
na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors
no = na * (nc + 5) # number of outputs = anchors * (classes + 5)
layers, save, c2 = [], [], ch[-1] # layers, savelist, ch out
for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']): # from, number, module, args
m_str = m
m = eval(m) if isinstance(m, str) else m # eval strings
for j, a in enumerate(args):
try:
args[j] = eval(a) if isinstance(a, str) else a # eval strings
except NameError:
pass
n = max(round(n * gd), 1) if n > 1 else n # depth gain
if m in [nn.Conv2d, Conv, Bottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, BottleneckCSP, C3]:
c1, c2 = ch[f], args[0]
c2 = make_divisible(c2 * gw, 8) if c2 != no else c2
args = [c1, c2, *args[1:]]
if m in [BottleneckCSP, C3]:
args.insert(2, n)
n = 1
elif m is nn.BatchNorm2d:
args = [ch[f]]
elif m is Concat:
c2 = sum(ch[-1 if x == -1 else x + 1] for x in f)
elif m is Detect:
args.append([ch[x + 1] for x in f])
if isinstance(args[1], int): # number of anchors
args[1] = [list(range(args[1] * 2))] * len(f)
args.append(imgsz)
else:
c2 = ch[f]
tf_m = eval('TF' + m_str.replace('nn.', ''))
m_ = keras.Sequential([tf_m(*args, w=model.model[i][j]) for j in range(n)]) if n > 1 \
else tf_m(*args, w=model.model[i]) # module
torch_m_ = nn.Sequential(*(m(*args) for _ in range(n))) if n > 1 else m(*args) # module
t = str(m)[8:-2].replace('__main__.', '') # module type
np = sum(x.numel() for x in torch_m_.parameters()) # number params
m_.i, m_.f, m_.type, m_.np = i, f, t, np # attach index, 'from' index, type, number params
LOGGER.info(f'{i:>3}{str(f):>18}{str(n):>3}{np:>10} {t:<40}{str(args):<30}') # print
save.extend(x % i for x in ([f] if isinstance(f, int) else f) if x != -1) # append to savelist
layers.append(m_)
ch.append(c2)
return keras.Sequential(layers), sorted(save)
class TFModel:
def __init__(self, cfg='yolov3.yaml', ch=3, nc=None, model=None, imgsz=(640, 640)): # model, channels, classes
super().__init__()
if isinstance(cfg, dict):
self.yaml = cfg # model dict
else: # is *.yaml
import yaml # for torch hub
self.yaml_file = Path(cfg).name
with open(cfg) as f:
self.yaml = yaml.load(f, Loader=yaml.FullLoader) # model dict
# Define model
if nc and nc != self.yaml['nc']:
LOGGER.info(f"Overriding {cfg} nc={self.yaml['nc']} with nc={nc}")
self.yaml['nc'] = nc # override yaml value
self.model, self.savelist = parse_model(deepcopy(self.yaml), ch=[ch], model=model, imgsz=imgsz)
def predict(self, inputs, tf_nms=False, agnostic_nms=False, topk_per_class=100, topk_all=100, iou_thres=0.45,
conf_thres=0.25):
y = [] # outputs
x = inputs
for i, m in enumerate(self.model.layers):
if m.f != -1: # if not from previous layer
x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers
x = m(x) # run
y.append(x if m.i in self.savelist else None) # save output
# Add TensorFlow NMS
if tf_nms:
boxes = self._xywh2xyxy(x[0][..., :4])
probs = x[0][:, :, 4:5]
classes = x[0][:, :, 5:]
scores = probs * classes
if agnostic_nms:
nms = AgnosticNMS()((boxes, classes, scores), topk_all, iou_thres, conf_thres)
return nms, x[1]
else:
boxes = tf.expand_dims(boxes, 2)
nms = tf.image.combined_non_max_suppression(
boxes, scores, topk_per_class, topk_all, iou_thres, conf_thres, clip_boxes=False)
return nms, x[1]
return x[0] # output only first tensor [1,6300,85] = [xywh, conf, class0, class1, ...]
# x = x[0][0] # [x(1,6300,85), ...] to x(6300,85)
# xywh = x[..., :4] # x(6300,4) boxes
# conf = x[..., 4:5] # x(6300,1) confidences
# cls = tf.reshape(tf.cast(tf.argmax(x[..., 5:], axis=1), tf.float32), (-1, 1)) # x(6300,1) classes
# return tf.concat([conf, cls, xywh], 1)
@staticmethod
def _xywh2xyxy(xywh):
# Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right
x, y, w, h = tf.split(xywh, num_or_size_splits=4, axis=-1)
return tf.concat([x - w / 2, y - h / 2, x + w / 2, y + h / 2], axis=-1)
class AgnosticNMS(keras.layers.Layer):
# TF Agnostic NMS
def call(self, input, topk_all, iou_thres, conf_thres):
# wrap map_fn to avoid TypeSpec related error https://stackoverflow.com/a/65809989/3036450
return tf.map_fn(lambda x: self._nms(x, topk_all, iou_thres, conf_thres), input,
fn_output_signature=(tf.float32, tf.float32, tf.float32, tf.int32),
name='agnostic_nms')
@staticmethod
def _nms(x, topk_all=100, iou_thres=0.45, conf_thres=0.25): # agnostic NMS
boxes, classes, scores = x
class_inds = tf.cast(tf.argmax(classes, axis=-1), tf.float32)
scores_inp = tf.reduce_max(scores, -1)
selected_inds = tf.image.non_max_suppression(
boxes, scores_inp, max_output_size=topk_all, iou_threshold=iou_thres, score_threshold=conf_thres)
selected_boxes = tf.gather(boxes, selected_inds)
padded_boxes = tf.pad(selected_boxes,
paddings=[[0, topk_all - tf.shape(selected_boxes)[0]], [0, 0]],
mode="CONSTANT", constant_values=0.0)
selected_scores = tf.gather(scores_inp, selected_inds)
padded_scores = tf.pad(selected_scores,
paddings=[[0, topk_all - tf.shape(selected_boxes)[0]]],
mode="CONSTANT", constant_values=-1.0)
selected_classes = tf.gather(class_inds, selected_inds)
padded_classes = tf.pad(selected_classes,
paddings=[[0, topk_all - tf.shape(selected_boxes)[0]]],
mode="CONSTANT", constant_values=-1.0)
valid_detections = tf.shape(selected_inds)[0]
return padded_boxes, padded_scores, padded_classes, valid_detections
def representative_dataset_gen(dataset, ncalib=100):
# Representative dataset generator for use with converter.representative_dataset, returns a generator of np arrays
for n, (path, img, im0s, vid_cap, string) in enumerate(dataset):
input = np.transpose(img, [1, 2, 0])
input = np.expand_dims(input, axis=0).astype(np.float32)
input /= 255
yield [input]
if n >= ncalib:
break
def run(weights=ROOT / 'yolov3.pt', # weights path
imgsz=(640, 640), # inference size h,w
batch_size=1, # batch size
dynamic=False, # dynamic batch size
):
# PyTorch model
im = torch.zeros((batch_size, 3, *imgsz)) # BCHW image
model = attempt_load(weights, map_location=torch.device('cpu'), inplace=True, fuse=False)
y = model(im) # inference
model.info()
# TensorFlow model
im = tf.zeros((batch_size, *imgsz, 3)) # BHWC image
tf_model = TFModel(cfg=model.yaml, model=model, nc=model.nc, imgsz=imgsz)
y = tf_model.predict(im) # inference
# Keras model
im = keras.Input(shape=(*imgsz, 3), batch_size=None if dynamic else batch_size)
keras_model = keras.Model(inputs=im, outputs=tf_model.predict(im))
keras_model.summary()
LOGGER.info('PyTorch, TensorFlow and Keras models successfully verified.\nUse export.py for TF model export.')
def parse_opt():
parser = argparse.ArgumentParser()
parser.add_argument('--weights', type=str, default=ROOT / 'yolov3.pt', help='weights path')
parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w')
parser.add_argument('--batch-size', type=int, default=1, help='batch size')
parser.add_argument('--dynamic', action='store_true', help='dynamic batch size')
opt = parser.parse_args()
opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand
print_args(FILE.stem, opt)
return opt
def main(opt):
run(**vars(opt))
if __name__ == "__main__":
opt = parse_opt()
main(opt)
models/yolo.py
+116 -88
View File
@@ -1,27 +1,32 @@
"""YOLOv3-specific modules
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
"""
YOLO-specific modules
Usage:
$ python path/to/models/yolo.py --cfg yolov3.yaml
"""
import argparse
import logging
import sys
from copy import deepcopy
from pathlib import Path
sys.path.append(Path(__file__).parent.parent.absolute().__str__()) # to run '$ python *.py' files in subdirectories
logger = logging.getLogger(__name__)
FILE = Path(__file__).resolve()
ROOT = FILE.parents[1] # root directory
if str(ROOT) not in sys.path:
sys.path.append(str(ROOT)) # add ROOT to PATH
# ROOT = ROOT.relative_to(Path.cwd()) # relative
from models.common import *
from models.experimental import *
from utils.autoanchor import check_anchor_order
from utils.general import make_divisible, check_file, set_logging
from utils.torch_utils import time_synchronized, fuse_conv_and_bn, model_info, scale_img, initialize_weights, \
select_device, copy_attr
from utils.general import LOGGER, check_version, check_yaml, make_divisible, print_args
from utils.plots import feature_visualization
from utils.torch_utils import (copy_attr, fuse_conv_and_bn, initialize_weights, model_info, scale_img, select_device,
time_sync)
try:
import thop # for FLOPS computation
import thop # for FLOPs computation
except ImportError:
thop = None
@@ -31,20 +36,18 @@ class Detect(nn.Module):
onnx_dynamic = False # ONNX export parameter
def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer
super(Detect, self).__init__()
super().__init__()
self.nc = nc # number of classes
self.no = nc + 5 # number of outputs per anchor
self.nl = len(anchors) # number of detection layers
self.na = len(anchors[0]) // 2 # number of anchors
self.grid = [torch.zeros(1)] * self.nl # init grid
a = torch.tensor(anchors).float().view(self.nl, -1, 2)
self.register_buffer('anchors', a) # shape(nl,na,2)
self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2)) # shape(nl,1,na,1,1,2)
self.anchor_grid = [torch.zeros(1)] * self.nl # init anchor grid
self.register_buffer('anchors', torch.tensor(anchors).float().view(self.nl, -1, 2)) # shape(nl,na,2)
self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv
self.inplace = inplace # use in-place ops (e.g. slice assignment)
def forward(self, x):
# x = x.copy() # for profiling
z = [] # inference output
for i in range(self.nl):
x[i] = self.m[i](x[i]) # conv
@@ -52,50 +55,55 @@ class Detect(nn.Module):
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
if not self.training: # inference
if self.grid[i].shape[2:4] != x[i].shape[2:4] or self.onnx_dynamic:
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
if self.onnx_dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]:
self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i)
y = x[i].sigmoid()
if self.inplace:
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy
y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953
xy = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i].view(1, self.na, 1, 1, 2) # wh
else: # for on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953
xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy
wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
y = torch.cat((xy, wh, y[..., 4:]), -1)
z.append(y.view(bs, -1, self.no))
return x if self.training else (torch.cat(z, 1), x)
@staticmethod
def _make_grid(nx=20, ny=20):
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
def _make_grid(self, nx=20, ny=20, i=0):
d = self.anchors[i].device
if check_version(torch.__version__, '1.10.0'): # torch>=1.10.0 meshgrid workaround for torch>=0.7 compatibility
yv, xv = torch.meshgrid([torch.arange(ny).to(d), torch.arange(nx).to(d)], indexing='ij')
else:
yv, xv = torch.meshgrid([torch.arange(ny).to(d), torch.arange(nx).to(d)])
grid = torch.stack((xv, yv), 2).expand((1, self.na, ny, nx, 2)).float()
anchor_grid = (self.anchors[i].clone() * self.stride[i]) \
.view((1, self.na, 1, 1, 2)).expand((1, self.na, ny, nx, 2)).float()
return grid, anchor_grid
class Model(nn.Module):
def __init__(self, cfg='yolov3.yaml', ch=3, nc=None, anchors=None): # model, input channels, number of classes
super(Model, self).__init__()
super().__init__()
if isinstance(cfg, dict):
self.yaml = cfg # model dict
else: # is *.yaml
import yaml # for torch hub
self.yaml_file = Path(cfg).name
with open(cfg) as f:
with open(cfg, encoding='ascii', errors='ignore') as f:
self.yaml = yaml.safe_load(f) # model dict
# Define model
ch = self.yaml['ch'] = self.yaml.get('ch', ch) # input channels
if nc and nc != self.yaml['nc']:
logger.info(f"Overriding model.yaml nc={self.yaml['nc']} with nc={nc}")
LOGGER.info(f"Overriding model.yaml nc={self.yaml['nc']} with nc={nc}")
self.yaml['nc'] = nc # override yaml value
if anchors:
logger.info(f'Overriding model.yaml anchors with anchors={anchors}')
LOGGER.info(f'Overriding model.yaml anchors with anchors={anchors}')
self.yaml['anchors'] = round(anchors) # override yaml value
self.model, self.save = parse_model(deepcopy(self.yaml), ch=[ch]) # model, savelist
self.names = [str(i) for i in range(self.yaml['nc'])] # default names
self.inplace = self.yaml.get('inplace', True)
# logger.info([x.shape for x in self.forward(torch.zeros(1, ch, 64, 64))])
# Build strides, anchors
m = self.model[-1] # Detect()
@@ -107,53 +115,42 @@ class Model(nn.Module):
check_anchor_order(m)
self.stride = m.stride
self._initialize_biases() # only run once
# logger.info('Strides: %s' % m.stride.tolist())
# Init weights, biases
initialize_weights(self)
self.info()
logger.info('')
LOGGER.info('')
def forward(self, x, augment=False, profile=False):
def forward(self, x, augment=False, profile=False, visualize=False):
if augment:
return self.forward_augment(x) # augmented inference, None
else:
return self.forward_once(x, profile) # single-scale inference, train
return self._forward_augment(x) # augmented inference, None
return self._forward_once(x, profile, visualize) # single-scale inference, train
def forward_augment(self, x):
def _forward_augment(self, x):
img_size = x.shape[-2:] # height, width
s = [1, 0.83, 0.67] # scales
f = [None, 3, None] # flips (2-ud, 3-lr)
y = [] # outputs
for si, fi in zip(s, f):
xi = scale_img(x.flip(fi) if fi else x, si, gs=int(self.stride.max()))
yi = self.forward_once(xi)[0] # forward
yi = self._forward_once(xi)[0] # forward
# cv2.imwrite(f'img_{si}.jpg', 255 * xi[0].cpu().numpy().transpose((1, 2, 0))[:, :, ::-1]) # save
yi = self._descale_pred(yi, fi, si, img_size)
y.append(yi)
y = self._clip_augmented(y) # clip augmented tails
return torch.cat(y, 1), None # augmented inference, train
def forward_once(self, x, profile=False):
def _forward_once(self, x, profile=False, visualize=False):
y, dt = [], [] # outputs
for m in self.model:
if m.f != -1: # if not from previous layer
x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers
if profile:
o = thop.profile(m, inputs=(x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPS
t = time_synchronized()
for _ in range(10):
_ = m(x)
dt.append((time_synchronized() - t) * 100)
if m == self.model[0]:
logger.info(f"{'time (ms)':>10s} {'GFLOPS':>10s} {'params':>10s} {'module'}")
logger.info(f'{dt[-1]:10.2f} {o:10.2f} {m.np:10.0f} {m.type}')
self._profile_one_layer(m, x, dt)
x = m(x) # run
y.append(x if m.i in self.save else None) # save output
if profile:
logger.info('%.1fms total' % sum(dt))
if visualize:
feature_visualization(x, m.type, m.i, save_dir=visualize)
return x
def _descale_pred(self, p, flips, scale, img_size):
@@ -173,6 +170,30 @@ class Model(nn.Module):
p = torch.cat((x, y, wh, p[..., 4:]), -1)
return p
def _clip_augmented(self, y):
# Clip augmented inference tails
nl = self.model[-1].nl # number of detection layers (P3-P5)
g = sum(4 ** x for x in range(nl)) # grid points
e = 1 # exclude layer count
i = (y[0].shape[1] // g) * sum(4 ** x for x in range(e)) # indices
y[0] = y[0][:, :-i] # large
i = (y[-1].shape[1] // g) * sum(4 ** (nl - 1 - x) for x in range(e)) # indices
y[-1] = y[-1][:, i:] # small
return y
def _profile_one_layer(self, m, x, dt):
c = isinstance(m, Detect) # is final layer, copy input as inplace fix
o = thop.profile(m, inputs=(x.copy() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPs
t = time_sync()
for _ in range(10):
m(x.copy() if c else x)
dt.append((time_sync() - t) * 100)
if m == self.model[0]:
LOGGER.info(f"{'time (ms)':>10s} {'GFLOPs':>10s} {'params':>10s} {'module'}")
LOGGER.info(f'{dt[-1]:10.2f} {o:10.2f} {m.np:10.0f} {m.type}')
if c:
LOGGER.info(f"{sum(dt):10.2f} {'-':>10s} {'-':>10s} Total")
def _initialize_biases(self, cf=None): # initialize biases into Detect(), cf is class frequency
# https://arxiv.org/abs/1708.02002 section 3.3
# cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1.
@@ -180,47 +201,33 @@ class Model(nn.Module):
for mi, s in zip(m.m, m.stride): # from
b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85)
b.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image)
b.data[:, 5:] += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum()) # cls
b.data[:, 5:] += math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum()) # cls
mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True)
def _print_biases(self):
m = self.model[-1] # Detect() module
for mi in m.m: # from
b = mi.bias.detach().view(m.na, -1).T # conv.bias(255) to (3,85)
logger.info(
LOGGER.info(
('%6g Conv2d.bias:' + '%10.3g' * 6) % (mi.weight.shape[1], *b[:5].mean(1).tolist(), b[5:].mean()))
# def _print_weights(self):
# for m in self.model.modules():
# if type(m) is Bottleneck:
# logger.info('%10.3g' % (m.w.detach().sigmoid() * 2)) # shortcut weights
# LOGGER.info('%10.3g' % (m.w.detach().sigmoid() * 2)) # shortcut weights
def fuse(self): # fuse model Conv2d() + BatchNorm2d() layers
logger.info('Fusing layers... ')
LOGGER.info('Fusing layers... ')
for m in self.model.modules():
if type(m) is Conv and hasattr(m, 'bn'):
if isinstance(m, (Conv, DWConv)) and hasattr(m, 'bn'):
m.conv = fuse_conv_and_bn(m.conv, m.bn) # update conv
delattr(m, 'bn') # remove batchnorm
m.forward = m.fuseforward # update forward
m.forward = m.forward_fuse # update forward
self.info()
return self
def nms(self, mode=True): # add or remove NMS module
present = type(self.model[-1]) is NMS # last layer is NMS
if mode and not present:
logger.info('Adding NMS... ')
m = NMS() # module
m.f = -1 # from
m.i = self.model[-1].i + 1 # index
self.model.add_module(name='%s' % m.i, module=m) # add
self.eval()
elif not mode and present:
logger.info('Removing NMS... ')
self.model = self.model[:-1] # remove
return self
def autoshape(self): # add AutoShape module
logger.info('Adding AutoShape... ')
LOGGER.info('Adding AutoShape... ')
m = AutoShape(self) # wrap model
copy_attr(m, self, include=('yaml', 'nc', 'hyp', 'names', 'stride'), exclude=()) # copy attributes
return m
@@ -228,9 +235,20 @@ class Model(nn.Module):
def info(self, verbose=False, img_size=640): # print model information
model_info(self, verbose, img_size)
def _apply(self, fn):
# Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers
self = super()._apply(fn)
m = self.model[-1] # Detect()
if isinstance(m, Detect):
m.stride = fn(m.stride)
m.grid = list(map(fn, m.grid))
if isinstance(m.anchor_grid, list):
m.anchor_grid = list(map(fn, m.anchor_grid))
return self
def parse_model(d, ch): # model_dict, input_channels(3)
logger.info('\n%3s%18s%3s%10s %-40s%-30s' % ('', 'from', 'n', 'params', 'module', 'arguments'))
LOGGER.info(f"\n{'':>3}{'from':>18}{'n':>3}{'params':>10} {'module':<40}{'arguments':<30}")
anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple']
na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors
no = na * (nc + 5) # number of outputs = anchors * (classes + 5)
@@ -241,24 +259,24 @@ def parse_model(d, ch): # model_dict, input_channels(3)
for j, a in enumerate(args):
try:
args[j] = eval(a) if isinstance(a, str) else a # eval strings
except:
except NameError:
pass
n = max(round(n * gd), 1) if n > 1 else n # depth gain
if m in [Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, DWConv, MixConv2d, Focus, CrossConv, BottleneckCSP,
C3, C3TR]:
n = n_ = max(round(n * gd), 1) if n > 1 else n # depth gain
if m in [Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv,
BottleneckCSP, C3, C3TR, C3SPP, C3Ghost]:
c1, c2 = ch[f], args[0]
if c2 != no: # if not output
c2 = make_divisible(c2 * gw, 8)
args = [c1, c2, *args[1:]]
if m in [BottleneckCSP, C3, C3TR]:
if m in [BottleneckCSP, C3, C3TR, C3Ghost]:
args.insert(2, n) # number of repeats
n = 1
elif m is nn.BatchNorm2d:
args = [ch[f]]
elif m is Concat:
c2 = sum([ch[x] for x in f])
c2 = sum(ch[x] for x in f)
elif m is Detect:
args.append([ch[x] for x in f])
if isinstance(args[1], int): # number of anchors
@@ -270,11 +288,11 @@ def parse_model(d, ch): # model_dict, input_channels(3)
else:
c2 = ch[f]
m_ = nn.Sequential(*[m(*args) for _ in range(n)]) if n > 1 else m(*args) # module
m_ = nn.Sequential(*(m(*args) for _ in range(n))) if n > 1 else m(*args) # module
t = str(m)[8:-2].replace('__main__.', '') # module type
np = sum([x.numel() for x in m_.parameters()]) # number params
np = sum(x.numel() for x in m_.parameters()) # number params
m_.i, m_.f, m_.type, m_.np = i, f, t, np # attach index, 'from' index, type, number params
logger.info('%3s%18s%3s%10.0f %-40s%-30s' % (i, f, n, np, t, args)) # print
LOGGER.info(f'{i:>3}{str(f):>18}{n_:>3}{np:10.0f} {t:<40}{str(args):<30}') # print
save.extend(x % i for x in ([f] if isinstance(f, int) else f) if x != -1) # append to savelist
layers.append(m_)
if i == 0:
@@ -285,11 +303,13 @@ def parse_model(d, ch): # model_dict, input_channels(3)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--cfg', type=str, default='yolov3.yaml', help='model.yaml')
parser.add_argument('--cfg', type=str, default='yolov3yaml', help='model.yaml')
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
parser.add_argument('--profile', action='store_true', help='profile model speed')
parser.add_argument('--test', action='store_true', help='test all yolo*.yaml')
opt = parser.parse_args()
opt.cfg = check_file(opt.cfg) # check file
set_logging()
opt.cfg = check_yaml(opt.cfg) # check YAML
print_args(FILE.stem, opt)
device = select_device(opt.device)
# Create model
@@ -297,12 +317,20 @@ if __name__ == '__main__':
model.train()
# Profile
# img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 320, 320).to(device)
# y = model(img, profile=True)
if opt.profile:
img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 640, 640).to(device)
y = model(img, profile=True)
# Test all models
if opt.test:
for cfg in Path(ROOT / 'models').rglob('yolo*.yaml'):
try:
_ = Model(cfg)
except Exception as e:
print(f'Error in {cfg}: {e}')
# Tensorboard (not working https://github.com/ultralytics/yolov5/issues/2898)
# from torch.utils.tensorboard import SummaryWriter
# tb_writer = SummaryWriter('.')
# logger.info("Run 'tensorboard --logdir=models' to view tensorboard at http://localhost:6006/")
# LOGGER.info("Run 'tensorboard --logdir=models' to view tensorboard at http://localhost:6006/")
# tb_writer.add_graph(torch.jit.trace(model, img, strict=False), []) # add model graph
# tb_writer.add_image('test', img[0], dataformats='CWH') # add model to tensorboard
models/yolov3-spp.yaml
+3 -3
View File
@@ -1,9 +1,9 @@
# parameters
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
# Parameters
nc: 80 # number of classes
depth_multiple: 1.0 # model depth multiple
width_multiple: 1.0 # layer channel multiple
# anchors
anchors:
- [10,13, 16,30, 33,23] # P3/8
- [30,61, 62,45, 59,119] # P4/16
models/yolov3-tiny.yaml
+3 -3
View File
@@ -1,9 +1,9 @@
# parameters
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
# Parameters
nc: 80 # number of classes
depth_multiple: 1.0 # model depth multiple
width_multiple: 1.0 # layer channel multiple
# anchors
anchors:
- [10,14, 23,27, 37,58] # P4/16
- [81,82, 135,169, 344,319] # P5/32
models/yolov3.yaml
+3 -3
View File
@@ -1,9 +1,9 @@
# parameters
# YOLOv3 🚀 by Ultralytics, GPL-3.0 license
# Parameters
nc: 80 # number of classes
depth_multiple: 1.0 # model depth multiple
width_multiple: 1.0 # layer channel multiple
# anchors
anchors:
- [10,13, 16,30, 33,23] # P3/8
- [30,61, 62,45, 59,119] # P4/16