步态识别-基于毫米波雷达的行人步态识别算法实现-基于Tensorflow-附项目源码-优质项目分享.zip

# Gait Based User Recognition from mmWave Radar Data ## Directory Structure - `models/`: Deep learning models/experiments - `resnet.py`: Vanilla Resnet base model imported by some models - `resnet_amca.py`: Resnet model with Constrictive Annular Loss(metric learning method) imported by some models - `utils.py`: Arbitrary methods used by most model scripts (loss functions, data preprocessing, confusion matrices, ...) - `preprocess/`: Script to generate spectrograms from raw data collected from radar - `tools/`: Arbitrary tools to plot data, list dataset files, generate embeddings from pretrained models - `mmwave_studio_scripts/`: Scripts used to collect radar data from [TI mmWave Studio](https://www.ti.com/tool/MMWAVE-STUDIO) along with video/depth data from [Microsoft Kinect Azure](https://azure.microsoft.com/en-us/services/kinect-dk/) camera. Use `*3d` files to record camera data along with radar data. ## Tripod mount for TI mmWave radar and cameras You can 3d print [this](https://www.thingiverse.com/thing:3780598) tripod mount for the TI radar and camera.