zjc
/
yolov5
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Merge remote-tracking branch 'origin/master'

This commit is contained in:
Glenn Jocher 2020-07-13 10:53:43 -07:00
parent 094079bd26 e169edfcf5
commit ea34f848a6
9 changed files with 55 additions and 38 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
README.md
View File

@ -21,7 +21,7 @@ This repository represents Ultralytics open-source research into future object d
| [YOLOv5m](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | 43.4 | 43.4 | 62.4 | 3.0ms | 333 || 21.8M | 39.4B | [YOLOv5m](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | 43.4 | 43.4 | 62.4 | 3.0ms | 333 || 21.8M | 39.4B
| [YOLOv5l](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | 46.6 | 46.7 | 65.4 | 3.9ms | 256 || 47.8M | 88.1B | [YOLOv5l](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | 46.6 | 46.7 | 65.4 | 3.9ms | 256 || 47.8M | 88.1B
| [YOLOv5x](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | **48.4** | **48.4** | **66.9** | 6.1ms | 164 || 89.0M | 166.4B | [YOLOv5x](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | **48.4** | **48.4** | **66.9** | 6.1ms | 164 || 89.0M | 166.4B
| [YOLOv3-SPP](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | 45.6 | 45.5 | 65.2 | 4.5ms | 222 || 63.0M | 118.0B | [YOLOv3-SPP](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) | 45.6 | 45.5 | 65.2 | 4.5ms | 222 || 63.0M | 118.0B
** AP<sup>test</sup> denotes COCO [test-dev2017](http://cocodataset.org/#upload) server results, all other AP results in the table denote val2017 accuracy. ** AP<sup>test</sup> denotes COCO [test-dev2017](http://cocodataset.org/#upload) server results, all other AP results in the table denote val2017 accuracy.
@ -54,10 +54,11 @@ $ pip install -U -r requirements.txt
Inference can be run on most common media formats. Model [checkpoints](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) are downloaded automatically if available. Results are saved to `./inference/output`. Inference can be run on most common media formats. Model [checkpoints](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J) are downloaded automatically if available. Results are saved to `./inference/output`.
```bash ```bash
$ python detect.py --source file.jpg # image $ python detect.py --source 0 # webcam
file.jpg # image
file.mp4 # video file.mp4 # video
./dir # directory path/ # directory
0 # webcam path/*.jpg # glob
rtsp://170.93.143.139/rtplive/470011e600ef003a004ee33696235daa # rtsp stream rtsp://170.93.143.139/rtplive/470011e600ef003a004ee33696235daa # rtsp stream
http://112.50.243.8/PLTV/88888888/224/3221225900/1.m3u8 # http stream http://112.50.243.8/PLTV/88888888/224/3221225900/1.m3u8 # http stream
``` ```

6
data/coco.yaml
View File

@ -1,13 +1,13 @@
# COCO 2017 dataset http://cocodataset.org # COCO 2017 dataset http://cocodataset.org
# Download command: bash yolov5/data/get_coco2017.sh # Download command: bash yolov5/data/get_coco2017.sh
# Train command: python train.py --data ./data/coco.yaml # Train command: python train.py --data coco.yaml
# Dataset should be placed next to yolov5 folder: # Default dataset location is next to /yolov5:
# /parent_folder # /parent_folder
# /coco # /coco
# /yolov5 # /yolov5
# train and val datasets (image directory or *.txt file with image paths) # train and val data as 1) directory: path/images/, 2) file: path/images.txt, or 3) list: [path1/images/, path2/images/]
train: ../coco/train2017.txt # 118k images train: ../coco/train2017.txt # 118k images
val: ../coco/val2017.txt # 5k images val: ../coco/val2017.txt # 5k images
test: ../coco/test-dev2017.txt # 20k images for submission to https://competitions.codalab.org/competitions/20794 test: ../coco/test-dev2017.txt # 20k images for submission to https://competitions.codalab.org/competitions/20794

12
data/coco128.yaml
View File

@ -1,15 +1,15 @@
# COCO 2017 dataset http://cocodataset.org - first 128 training images # COCO 2017 dataset http://cocodataset.org - first 128 training images
# Download command: python -c "from yolov5.utils.google_utils import gdrive_download; gdrive_download('1n_oKgR81BJtqk75b00eAjdv03qVCQn2f','coco128.zip')" # Download command: python -c "from yolov5.utils.google_utils import *; gdrive_download('1n_oKgR81BJtqk75b00eAjdv03qVCQn2f', 'coco128.zip')"
# Train command: python train.py --data ./data/coco128.yaml # Train command: python train.py --data coco128.yaml
# Dataset should be placed next to yolov5 folder: # Default dataset location is next to /yolov5:
# /parent_folder # /parent_folder
# /coco128 # /coco128
# /yolov5 # /yolov5
# train and val datasets (image directory or *.txt file with image paths) # train and val data as 1) directory: path/images/, 2) file: path/images.txt, or 3) list: [path1/images/, path2/images/]
train: ../coco128/images/train2017/ train: ../coco128/images/train2017/ # 128 images
val: ../coco128/images/train2017/ val: ../coco128/images/train2017/ # 128 images
# number of classes # number of classes
nc: 80 nc: 80

5
data/get_coco2017.sh
View File

@ -1,12 +1,13 @@
#!/bin/bash #!/bin/bash
# COCO 2017 dataset http://cocodataset.org # COCO 2017 dataset http://cocodataset.org
# Download command: bash yolov5/data/get_coco2017.sh # Download command: bash yolov5/data/get_coco2017.sh
# Train command: python train.py --data ./data/coco.yaml # Train command: python train.py --data coco.yaml
# Dataset should be placed next to yolov5 folder: # Default dataset location is next to /yolov5:
# /parent_folder # /parent_folder
# /coco # /coco
# /yolov5 # /yolov5
# Download labels from Google Drive, accepting presented query # Download labels from Google Drive, accepting presented query
filename="coco2017labels.zip" filename="coco2017labels.zip"
fileid="1cXZR_ckHki6nddOmcysCuuJFM--T-Q6L" fileid="1cXZR_ckHki6nddOmcysCuuJFM--T-Q6L"

3
data/get_voc.sh
View File

@ -1,11 +1,12 @@
# PASCAL VOC dataset http://host.robots.ox.ac.uk/pascal/VOC/ # PASCAL VOC dataset http://host.robots.ox.ac.uk/pascal/VOC/
# Download command: bash ./data/get_voc.sh # Download command: bash ./data/get_voc.sh
# Train command: python train.py --data voc.yaml # Train command: python train.py --data voc.yaml
# Dataset should be placed next to yolov5 folder: # Default dataset location is next to /yolov5:
# /parent_folder # /parent_folder
# /VOC # /VOC
# /yolov5 # /yolov5
start=`date +%s` start=`date +%s`
# handle optional download dir # handle optional download dir

9
data/voc.yaml
View File

@ -1,14 +1,15 @@
# PASCAL VOC dataset http://host.robots.ox.ac.uk/pascal/VOC/ # PASCAL VOC dataset http://host.robots.ox.ac.uk/pascal/VOC/
# Download command: bash ./data/get_voc.sh # Download command: bash ./data/get_voc.sh
# Train command: python train.py --data voc.yaml # Train command: python train.py --data voc.yaml
# Dataset should be placed next to yolov5 folder: # Default dataset location is next to /yolov5:
# /parent_folder # /parent_folder
# /VOC # /VOC
# /yolov5 # /yolov5
# train and val datasets (image directory or *.txt file with image paths)
train: ../VOC/images/train/ # train and val data as 1) directory: path/images/, 2) file: path/images.txt, or 3) list: [path1/images/, path2/images/]
val: ../VOC/images/val/ train: ../VOC/images/train/ # 16551 images
val: ../VOC/images/val/ # 4952 images
# number of classes # number of classes
nc: 20 nc: 20

4
test.py
View File

@ -41,9 +41,9 @@ def test(data,
# model = nn.DataParallel(model) # model = nn.DataParallel(model)
# Half # Half
half = device.type != 'cpu' and torch.cuda.device_count() == 1 # half precision only supported on single-GPU half = device.type != 'cpu' # half precision only supported on CUDA
if half: if half:
model.half() # to FP16 model.half()
# Configure # Configure
model.eval() model.eval()

43
utils/datasets.py
View File

@ -68,35 +68,39 @@ def create_dataloader(path, imgsz, batch_size, stride, opt, hyp=None, augment=Fa
class LoadImages: # for inference class LoadImages: # for inference
def __init__(self, path, img_size=640): def __init__(self, path, img_size=640):
path = str(Path(path)) # os-agnostic p = str(Path(path)) # os-agnostic
files = [] p = os.path.abspath(p) # absolute path
if os.path.isdir(path): if '*' in p:
files = sorted(glob.glob(os.path.join(path, '*.*'))) files = sorted(glob.glob(p)) # glob
elif os.path.isfile(path): elif os.path.isdir(p):
files = [path] files = sorted(glob.glob(os.path.join(p, '*.*'))) # dir
elif os.path.isfile(p):
files = [p] # files
else:
raise Exception('ERROR: %s does not exist' % p)
images = [x for x in files if os.path.splitext(x)[-1].lower() in img_formats] images = [x for x in files if os.path.splitext(x)[-1].lower() in img_formats]
videos = [x for x in files if os.path.splitext(x)[-1].lower() in vid_formats] videos = [x for x in files if os.path.splitext(x)[-1].lower() in vid_formats]
nI, nV = len(images), len(videos) ni, nv = len(images), len(videos)
self.img_size = img_size self.img_size = img_size
self.files = images + videos self.files = images + videos
self.nF = nI + nV # number of files self.nf = ni + nv # number of files
self.video_flag = [False] * nI + [True] * nV self.video_flag = [False] * ni + [True] * nv
self.mode = 'images' self.mode = 'images'
if any(videos): if any(videos):
self.new_video(videos[0]) # new video self.new_video(videos[0]) # new video
else: else:
self.cap = None self.cap = None
assert self.nF > 0, 'No images or videos found in %s. Supported formats are:\nimages: %s\nvideos: %s' % \ assert self.nf > 0, 'No images or videos found in %s. Supported formats are:\nimages: %s\nvideos: %s' % \
(path, img_formats, vid_formats) (p, img_formats, vid_formats)
def __iter__(self): def __iter__(self):
self.count = 0 self.count = 0
return self return self
def __next__(self): def __next__(self):
if self.count == self.nF: if self.count == self.nf:
raise StopIteration raise StopIteration
path = self.files[self.count] path = self.files[self.count]
@ -107,7 +111,7 @@ class LoadImages: # for inference
if not ret_val: if not ret_val:
self.count += 1 self.count += 1
self.cap.release() self.cap.release()
if self.count == self.nF: # last video if self.count == self.nf: # last video
raise StopIteration raise StopIteration
else: else:
path = self.files[self.count] path = self.files[self.count]
@ -115,14 +119,14 @@ class LoadImages: # for inference
ret_val, img0 = self.cap.read() ret_val, img0 = self.cap.read()
self.frame += 1 self.frame += 1
print('video %g/%g (%g/%g) %s: ' % (self.count + 1, self.nF, self.frame, self.nframes, path), end='') print('video %g/%g (%g/%g) %s: ' % (self.count + 1, self.nf, self.frame, self.nframes, path), end='')
else: else:
# Read image # Read image
self.count += 1 self.count += 1
img0 = cv2.imread(path) # BGR img0 = cv2.imread(path) # BGR
assert img0 is not None, 'Image Not Found ' + path assert img0 is not None, 'Image Not Found ' + path
print('image %g/%g %s: ' % (self.count, self.nF, path), end='') print('image %g/%g %s: ' % (self.count, self.nf, path), end='')
# Padded resize # Padded resize
img = letterbox(img0, new_shape=self.img_size)[0] img = letterbox(img0, new_shape=self.img_size)[0]
@ -140,7 +144,7 @@ class LoadImages: # for inference
self.nframes = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT)) self.nframes = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT))
def __len__(self): def __len__(self):
return self.nF # number of files return self.nf # number of files
class LoadWebcam: # for inference class LoadWebcam: # for inference
@ -470,6 +474,13 @@ class LoadImagesAndLabels(Dataset): # for training/testing
img, labels = load_mosaic(self, index) img, labels = load_mosaic(self, index)
shapes = None shapes = None
# MixUp https://arxiv.org/pdf/1710.09412.pdf
# if random.random() < 0.5:
# img2, labels2 = load_mosaic(self, random.randint(0, len(self.labels) - 1))
# r = np.random.beta(0.3, 0.3) # mixup ratio, alpha=beta=0.3
# img = (img * r + img2 * (1 - r)).astype(np.uint8)
# labels = np.concatenate((labels, labels2), 0)
else: else:
# Load image # Load image
img, (h0, w0), (h, w) = load_image(self, index) img, (h0, w0), (h, w) = load_image(self, index)

2
utils/torch_utils.py
View File

@ -195,6 +195,8 @@ class ModelEMA:
def __init__(self, model, decay=0.9999, updates=0): def __init__(self, model, decay=0.9999, updates=0):
# Create EMA # Create EMA
self.ema = deepcopy(model.module if is_parallel(model) else model).eval() # FP32 EMA self.ema = deepcopy(model.module if is_parallel(model) else model).eval() # FP32 EMA
if next(model.parameters()).device.type != 'cpu':
self.ema.half() # FP16 EMA
self.updates = updates # number of EMA updates self.updates = updates # number of EMA updates
self.decay = lambda x: decay * (1 - math.exp(-x / 2000)) # decay exponential ramp (to help early epochs) self.decay = lambda x: decay * (1 - math.exp(-x / 2000)) # decay exponential ramp (to help early epochs)
for p in self.ema.parameters(): for p in self.ema.parameters():