Skip to content

Latest commit

 

History

History
45 lines (30 loc) · 1.63 KB

README.md

File metadata and controls

45 lines (30 loc) · 1.63 KB

Folder Structure

dataCollector/data:

  • coco.txt: Text file with several object classes in string form.
  • dataGenerator.py: Algorithm to expand our dataset.
  • main.py: Algorithm to count objects from IP webcam (Android only).
  • mySample.mp4: Small video sample for test usage.
  • resampled_data.csv: Data after expansion.
  • sample2.mp4: Another video sample for test usage.
  • test.py: Test file to see if IP webcam is working.
  • tracker.py: Tracker() class for object tracking (do NOT modify).
  • traffic_data_6_2023.csv: Data we collected from a real-time video.
  • yolov8s.pt: Pre-trained weights of YOLO model to recognize objects (do NOT delete or modify).

trafficPrediction: Use of our AI model

data:

  • simple_preprocessor.py: Preprocess data for our AI model.
  • train.csv: Training dataset.
  • test.csv: Testing dataset.

images:

  • data.png: Graph that shows the predicted data vs the actual data.
  • training_mae.png: Graph that shows training and validation error.

model:

  • cnn1D.py: Contains 3 functions (one for each model including CNN, RNN).
    • Hint: If you want to change the model, just change the function name in train_model and that's it.

weights:

  • trained_cnn.h5: Saved binary form of our model's trained weights to use them in prediction and avoid training every time.

  • model_architecture.png

  • predict_traffic.py: Algorithm to predict the traffic for 1 month.

  • train_model.py: Algorithm to train our model using preprocessed data of 2 months.

For directions on how to run and more explanations, please look inside the files.