Abstract:Multi-area visitor counting plays a critical role in museum management, which can help administrative staff better study visitor flows and hotspots, so that they can ensure the quality and safety of visits. Internet of Things (IoT) techniques facilitate efficient recording and understanding of visitors' spatial and temporal distribution in museums, and traditional visitor tracking applications use surveillance cameras or wireless connections with portable smart devices. However, these methods either involve privacy concerns or face the obstacle of getting natural behavioral data of all visitors. This article explores an IoT monitoring methodology in the field of museum studies, proposing a commodity WiFi-based head-counting framework that does not need the visitor to connect with any device. Our system analyzes the Channel State Information amplitude fluctuations at the fixed receiver caused when visitors cross the line-of-sight link. It enables multi-area visitor counting by achieving In-and-Out traffic detection at different sites with a convolutional neural network algorithm. The method also allows for a rough classification of visitor types based on body size, and an extra transfer module is presented to reduce training time for increasing sensing scenarios. Over 2,300 samples at five different sites were collected to test the usability. Experiment 1 implemented in three environments/deployments demonstrated that the proposed approach can be potentially implemented in variable sites of museums. It achieved high up to 95% and 99% accuracies for identifying the number and direction of people crossing, respectively. Experiment 2 sampled adults, children, and adult-child groups at a science museum and achieved approximately 89% classification accuracy of visitor types. Experiment 3 collected data for all cases in which up to three targets entered and exited simultaneously, and reached a recognition accuracy of around 88% for nine different cases. The potential and limitations for the practical application of wireless contactless sensing to cultural spaces are discussed.

Deep-Learning-Based Occupant Counting by Ambient RF Sensing

DeepCount: Crowd Counting with WiFi via Deep Learning

Adaptive Scheme for Crowd Counting Using Off-the-shelf Wireless Routers

3-D Indoor Device-Free Object Detection by Passive Radio Frequency Identification.

Novel Robust On-Line Indoor Occupancy Counting System Using mmWave Radar

Device-free crowd counting using multi-link WiFi CSI for occupant-driven energy management of HVAC systems

Indoor room Occupancy Counting based on LSTM and Environmental Sensor

People Counting with Carry-on RFID Tags.

Counting Human Objects Using Backscattered Radio Frequency Signals.

Harvesting Ambient RF for Presence Detection Through Deep Learning

High-accuracy Occupancy Counting at Crowded Entrances for Smart Buildings

Human Occupancy Detection via Passive Cognitive Radio

Improving RF-based device-free passive localization in cluttered indoor environments through probabilistic classification methods

Human Object Estimation Via Backscattered Radio Frequency Signal.

Building Occupancy Detection and Localization Using CCTV Camera and Deep Learning

Measuring and Monitoring Occupancy with an Rfid Based System for Demand-Driven Hvac Operations

Poster: Statistical Learning Strategies for RF-based Indoor Device-Free Passive Localization

Toward Multi-area Contactless Museum Visitor Counting with Commodity WiFi

RFID-Based Occupancy Detection Solution for Optimizing HVAC Energy Consumption

Scpl: Indoor Device-Free Multi-Subject Counting And Localization Using Radio Signal Strength

CD-HOC: Indoor Human Occupancy Counting using Carbon Dioxide Sensor Data