RateMe is a neural network that allows you to recognize gestures of thumb up and thumb down. The algorithm can be embedded in your project and automate the process of evaluation of something or someone.
- numpy
- opencv-python-inference-engine>=2021-10-10
Minimal working example:
import cv2
from rateme.utils import RateMe
net = RateMe()
img = cv2.imread('test_imgs/like.jpg')
label = net.predict(img)RateMe is based on tiny-YOLOv3 architecture.
It's accuracy of thumb up/down gesture recognition is calculated as mean average precision ([email protected]) = 0.851941, or 85.19%; average IoU = 73.89%
The neural network has been trained on ~3K images (taken from different angles photos of people showing their thumbs or not). Images were labeled using the labelImg program.
Return values: "like", "dislike", None
Full pipeline speed is 6-7 FPS on Intel(R) Core(TM) i5-4300M CPU @ 2.60GHz.
~100ms on frame grabbing
~100ms on neural network inference
cp -dr rateme create_package/
cd create_package
python setup.py sdist