Recognizing Activities of Daily Living (ADL) is crucial for applications such as elderly monitoring and home automation. Traditional methods often require separate systems for activity recognition and data annotation, making them complex and resource-intensive. In this paper, we present a unified approach that leverages WiFi technologies, specifically WiFi Channel State Information (CSI) and frequency-shift backscatter, to simultaneously achieve both ADL recognition and efficient data annotation. Our system utilizes WiFi-backscatter tags attached to individuals, enabling the automatic generation of pseudo-labels that annotate CSI data in real-time. This approach reduces the need for manual data labeling and model retraining, even as environmental conditions change. This simplifies the deployment and maintenance of ADL recognition systems. We validated our approach in a smart home environment on a university campus, specifically within a 3.5 m \(\times \) 2.6 m room. The experimental results suggest that our WiFi-based solution simplifies the process of ADL recognition and has the potential to make it more scalable and practical by integrating data annotation within the WiFi ecosystem.