Counting curls and making sure the elbows are tucked in using Yolov7 pose estimation and OpenCV.
Using Yolov5, DeepSort and Supervision to track the number of sacks loaded onto and off of a truck.
A web interface for real-time yolo inference using streamlit. It supports CPU and GPU inference, supports both images and videos and uploading your own custom models.
- Caches the model for faster inference on both CPU and GPU.
- Supports uploading model files (<200MB) and downloading models from URL (any size)
- Supports both images and videos.
- Supports both CPU and GPU inference.
- Supports Custom Classes and changing Confidence.
Labelling your custom datasets for object detection can take a huge time. So why don’t we use pretrained models to auto-label our objects for us?
In this article, I will explain how you can use pretrained models to speed up the labelling process for your next object detection project
