Security is always a matter of utmost importance. If security incidents can be detected early and stopped early, the result may be completely different. The core purpose of this article is very simple. It is to try to build a detection model for dangerous objects such as guns and knives based on the target detection model. The recognition system hopes to combat crimes based on artificial intelligence means, simply look at the renderings:
This kind of scene is relatively sensitive, so it is basically difficult to actually collect the data set by yourself to develop the model, so most of the data here comes from Huchao.com, mainly for practical analysis and use. A brief look at the data set is as follows :
The VOC format annotation file is as follows:
The content of the example annotation is as follows:
<annotation>
<folder>DATASET</folder>
<filename>images/0a53eb41-a382-4463-ad29-3e63bef6420d.jpg</filename>
<source>
<database>The DATASET Database</database>
<annotation>DATASET</annotation>
<image>DATASET</image>
</source>
<owner>
<name>YMGZS</name>
</owner>
<size>
<width>416</width>
<height>416</height>
<depth>3</depth>
</size>
<segmented>0</segmented>
<object>
<name>pistol</name>
<pose>Unspecified</pose>
<truncated>0</truncated>
<difficult>0</difficult>
<bndbox>
<xmin>3</xmin>
<ymin>7</ymin>
<xmax>403</xmax>
<ymax>395</ymax>
</bndbox>
</object>
</annotation>The YOLO format annotation file is as follows:
The content of the example annotation is as follows:
1 0.546875 0.492788 0.122596 0.466346
The core training configuration looks like this:
def parse_opt(known=False):
parser = argparse.ArgumentParser()
parser.add_argument('--weights', type=str, default='weights/yolov5n.pt', help='initial weights path')
parser.add_argument('--cfg', type=str, default='models/yolov5n.yaml', help='model.yaml path')
parser.add_argument('--data', type=str, default='data/self.yaml', help='dataset.yaml path')
parser.add_argument('--hyp', type=str, default='data/hyps/hyp.scratch-low.yaml', help='hyperparameters path')
parser.add_argument('--epochs', type=int, default=100, help='total training epochs')
parser.add_argument('--batch-size', type=int, default=8, help='total batch size for all GPUs, -1 for autobatch')
parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=416, help='train, val image size (pixels)')
parser.add_argument('--rect', action='store_true', help='rectangular training')
parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training')
parser.add_argument('--nosave', action='store_true', help='only save final checkpoint')
parser.add_argument('--noval', action='store_true', help='only validate final epoch')
parser.add_argument('--noautoanchor', action='store_true', help='disable AutoAnchor')
parser.add_argument('--noplots', action='store_true', help='save no plot files')
parser.add_argument('--evolve', type=int, nargs='?', const=300, help='evolve hyperparameters for x generations')
parser.add_argument('--bucket', type=str, default='', help='gsutil bucket')
parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"')
parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training')
parser.add_argument('--device', default='0', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%')
parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class')
parser.add_argument('--optimizer', type=str, choices=['SGD', 'Adam', 'AdamW'], default='SGD', help='optimizer')
parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode')
parser.add_argument('--workers', type=int, default=0, help='max dataloader workers (per RANK in DDP mode)')
parser.add_argument('--project', default='runs/train', help='save to project/name')
parser.add_argument('--name', default='yolov5n', help='save to project/name')
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
parser.add_argument('--quad', action='store_true', help='quad dataloader')
parser.add_argument('--cos-lr', action='store_true', help='cosine LR scheduler')
parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon')
parser.add_argument('--patience', type=int, default=100, help='EarlyStopping patience (epochs without improvement)')
parser.add_argument('--freeze', nargs='+', type=int, default=[0], help='Freeze layers: backbone=10, first3=0 1 2')
parser.add_argument('--save-period', type=int, default=-1, help='Save checkpoint every x epochs (disabled if < 1)')
parser.add_argument('--seed', type=int, default=0, help='Global training seed')
parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify')
You can see: I am using the n-series model here, the image size is 416x416, and the iterative calculation of 100 epochs is performed by default.
The result details are as follows:
The F1 value and PR curve are as follows:
The precision and recall curves are as follows:
The overall training visualization is as follows:
The batch calculation example is as follows:
You can try it if you are interested.
The confusion matrix is as follows:
The data visualization is as follows:
