Skip to content

tcnitt/TransfiNITTe-V2.0_Bosons

Repository files navigation

TransfiNITTe-V2.0_Bosons

Deep learning based occupants tracker

We implement a method to count the number of occupants in room, from the given video feed of the room. We use a YOLOv3 model as our backbone. The model detects humans in the room and we count the number of predictions in each frame to give out the final count.

Running locally

  • Download pretrained YOLOv3 weights and it's config files from the below link. https://pjreddie.com/darknet/yolo/
  • Install OpenCV
  • To run inference on an image, run the below command in your terminal: python inference.py --image <image_path> --config <yolo_config_path> --weights <yolo_weights_path> --classes <no_of_class>

Future improvements

  • Currently this model runs just using OpenCV and DNN modules, we can speed it up by using TFLite.
  • Using more latest object detection models like YOLOv5 or FRCNN.


Team Members

  • Shrikumaran PB
  • Kailash J
  • Akshat K
  • Hemangani N
  • Akesh M
  • Sathya Prakash

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •