/Segment-and-Track-Human-Using-Wireless-Signal

This study is processed in [11785 - Introduction to Deep Learning] of Carnegie Mellon University in 2022 spring semester. Segments human instances and tracks human identities in a unified fashion based on WiFi channel state information (CSI).

Primary LanguagePython

Segment-and-Track-Human-Using-Wireless-Signal

Human segmentation and tracking studied in [11785 - Introduction to Deep Learning] of Carnegie Mellon University in 2022 spring semester.

Preparation

1. Create the environment.

$ cd ~/Segment-and-Track-Human-Using-Wireless-Signal
$ sh environment.sh

2. Place the data and label in the path below.

Segment-and-Track-Human-Using-Wireless-Signal/    
└── dataset/
    └── data
          └── s1p0_1
          └── s1p0_2
          └── s1p0_3
                .
                .
          └── s1p2_9
          └── s1p2_10
    
    └── label
          └── s1p0_1
          └── s1p0_2
          └── s1p0_3
                .
                .
          └── s1p2_9
          └── s1p2_10

3. Run main.sh for train

$ cd ~/Segment-and-Track-Human-Using-Wireless-Signal
$ sh train.sh

4. Run test.sh for test

$ cd ~/Segment-and-Track-Human-Using-Wireless-Signal
$ sh test.sh