nkmhd3 / yolov3

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

YOLO v3 : Real Time Object Detection

You only look once (YOLO) is a state-of-the-art, real-time object detection system. On a Pascal Titan X it processes images at 30 FPS and has a mAP of 57.9% on COCO test-dev YOLO is a system for detecting objects on the Pascal VOC 2012 dataset. It can detect about 20 Pascal object classes.

But How?

The prior detection systems repurpose classifiers or localizers to perform detection. They apply the model to an image at multiple locations and scales. High scoring regions of the image are considered detections. Here we use a totally different approach. We apply a single neural network to the full image. This network divides the image into regions and predicts bounding boxes and probabilities for each region. These bounding boxes are weighted by the predicted probabilities.

Wanna know more?
https://bit.ly/2Cnv869

Requirements

Anaconda – Python 3.7 (Windows 10) https://www.anaconda.com/download/

Conda Env yolov3.yml https://github.com/astra-sv/yolov3/blob/master/yolov3.yml

 cd C:\yolo
 conda env create -f yolo.yml

If any errors popup, try installing from requirements.txt

pip install -r requirements.txt

CUDA Toolkit - V9.0 (Nvidia GPU GTX 1050 or higher) https://bit.ly/2PRLjke

Ensure that you have the latest nvidia graphics drivers install on your PC.

PyTorch

Change directory to a workplace where you want to download. Clone Yolo v3 Repo

 cd C:\yolotorch

Download test video (.mp4/.avi) Rename video to video_demo.mp4

  python video_demo.py --video video.mp4

To run demo on webcam

  python cam_demo.py

Optional

If you get this error while running the webcam

   File "cam_demo.py" , line 123, in <module>
   im_dim=im_dim.cuda
   NameError: name 'im_dim' is not defined

Try opening cam_demo.py and uncomment the line 119

About