Use W&B’s lightweight, interoperable tools to quickly track experiments, version and iterate on datasets, evaluate model performance, reproduce models, visualize results and spot regressions, and share findings with colleagues. 0. Command setup wandb #pip install wandb #wandb login start a new run import wandb wandb.init(project="my-test-project") Track metrics wandb.log({'accuracy': train_acc, 'loss': train_loss}) track hypermeters wandb.config.dropout = 0.2 1. experiment tracking wandb.init(): Initialize a new run at the top of your script.

1. Introduce a wireless home automation protocol; an alliance of companies; all z-ware hardware uses sigma chips; z-wave has been a closed protocol; devices and protocol managed by sigma and z-wave alliance; certified devices undergo a certification scheme to ensure compatibility; open source projects needed to rely on reverse engineering the protocol; approx 600 companies and 2200 certified devices; introduce in 2015 and backward compatible with the original standard; improved devices with lower battery comsumption, coupled with improved protocol functionality; data rate 100kb/s; lower power consumption for better battery life; 2.

密码学包括：密码编码学和密码分析学。 密码编码学：主要研究密码方案的设计，即寻找对信息编码的方法从而实现隐藏信息的一门学问。 密码分析学：主要是

RFID系统本身的安全问题可归纳为隐私和认证两个方面：在隐私方面主要是可追踪性问题，即如何防止攻击者对RFID标签进行任何形式的跟踪; 在认证

level: CCF_A Mobicom author: Jinsong Han(ZJU); Lei Yang(Hongkong); Fei Wang, Feng Lin(ZJU) date: 2020 keyword: wireless sensing, privacy, RFBP Liu, Jianwei, et al. “A Framework for Behavior Privacy Preserving in Radio Frequency Signal.” arXiv preprint arXiv:2004.04909 (2020). Paper: Behavior Privacy Summary first notice that users’ behavior privacy may be filched（偷窃） in RF signal-based applications, and then propose the concept of

level: CVPR author: Boqiang Xu (University of Chinese Academy of Sciences), Lingxiao He(AI Research of JD) date: 2020 keyword: Person re-identification Paper: Black Re-ID Summary propose the study of the Black Re-ID problem and establish the first Black-reID dataset; propose the head-shoulder adaptive attention network(HAA), which make use of the head-shoulder information to support person re-identification through the adaptive attention module, and can be integrated with the most current Re-ID