noparkee / ConcentrationLecture

Deep Learning on Human Concentration

Overview

figure 1. project overview

Contributors

• WoodoLee
• lawkelvin33
• noparkee
• ujos89

Flow

Step0. Environment Setting

ubuntu 20.04 gpu: GTX 1060 6GB

• main environment
  • conda 4.8.3
  • python 3.6.12
• harness gpu
  • cuda 10.0
  • cudnn 7.6.5
  • tensorflow 1.15
  • keras 2.3.1
• visualization
  • matplotlib 3.3.1
  • seaborn 0.11.1
• process data
  • pandas 1.1.1
  • sklearn 0.23.2

Step1. Video data to Pickle (video2pickle.py)

$ python3 video2pickle.py --video [video_name] --savefile [file_name_to_save]

We used ildoonet/tf-pose-estimation to extract each body part informations save informations to pickle

Step2. Data Preprocessing (preprocessing.py)

$ python3 preprocessing.py --rawroot [raw_file_name]

In preprocessing.py...

• top: Nos, Lea, Ley, Rea, Rey
• mid: Nec, Lel, Lsh, Rel, Rsh
• Calculate the variations of the top and mid part per every 50 frames and add labels.

Step3. Merge prepared dataset & Shuffle

$ python3 build_trainset.py --name [person_initial] --index [index_number] 

merge dataset and shuffle to prevent biased labeled value

Step4. Training DNN

$ python3 run_dnn.py --file [name of pickle] --plot [graph_idx] --size [dataset_size] --epoch [number of epoch] 

• 1st hidden layer: dimension:8, activation: relu
• 2nd hidden layer: dimension:8, activation: relu
• output layer: sigmoid

Step5. Analysis

compared between size of data