With increasing attention to the construction of barrier-free environments, the “information inequity” faced by hearing-impaired individuals has become more pronounced, especially in scenarios such as lack of auditory alerts and difficulties in decoding social cues. Traditional hearing aids suffer from functional limitations, discomfort in wearing, and high costs, while single-modal sensory compensation solutions are often insufficient to meet complex information transmission needs. To address these challenges, this study proposes a wearable sensing device based on a dual-perception compensation mechanism. By constructing an interactive chain of “sound acquisition—semantic recognition—multimodal feedback,” the system transforms environmental sound signals into visual light effects and tactile vibrations in real time, enabling the identification and feedback of two key types of information: “safety alerts” and “social semantics.” The system employs lightweight AI models for acoustic classification and keyword recognition, and uses linear resonant actuators and LED flowing light effects to encode information content and spatial direction. It also supports integration with mobile apps to enhance communication efficiency. This research provides a new technical approach toward achieving “information equity,” offering broad social application prospects and valuable design insights.