Evaluating badminton performance often requires expert coaching, which is rarely accessible for amateur players. We present BadminSense, a smartwatch-based system for fine-grained badminton performance analysis using wearable sensing. Through interviews with experienced badminton players, we identified four system design requirements with three implementation insights that guide the development of BadminSense. We then collected a badminton strokes dataset on 12 experienced badminton amateurs and annotated it with fine-grained labels, including stroke type, expert-assessed stroke rating, and shuttle impact location. Built on this dataset, BadminSense segments and classifies strokes, predicts stroke quality, and estimates shuttle impact location using vibration signal from an off-the-shelf smartwatch. Our evaluations show that BadminSense achieves a stroke classification accuracy of 91.43%, an average quality rating error of 0.438, and an average impact location estimation error of 12.9%. A real-world usability study further demonstrates BadminSense’s potential to provide reliable and meaningful support for daily badminton practice.

Redirected Walking (RDW) enables navigation in large virtual worlds by manipulating a user's viewpoint, yet the perceptual thresholds governing these manipulations are often treated as fixed sensory limits. In this work, we propose the concept of Semantic Anchors: objects that stabilise self-motion perception through high-level cognitive priors. In a psychophysical experiment $(N=22)$, participants judged virtual rotations while viewing familiar or arbitrary objects of varying sizes and rotational reference frames. Results reveal that scene semantics do not merely modulate thresholds but fundamentally alter the integration of spatial cues. When a familiar object's canonical size was violated to create a specific size-congruency conflict, the window for imperceptible redirection expanded significantly, allowing rotation gains of up to 30% to remain imperceptible. Crucially, familiarity inverted the geometric dependence observed with arbitrary shapes: while geometric frames dictated sensitivity for abstract objects, semantic stability dominated for familiar ones. These findings demonstrate that RDW thresholds are dynamically co-constructed by sensory input and scene plausibility, offering a new principle for designing semantically informed locomotion systems that leverage cognitive expectations to expand explorable space.

There has been an increasing focus on haptic interfaces for virtual reality (VR), to support a high-quality touch experience. However, it is still challenging to haptically simulate the real-world walking experience in different fluid mediums. To tackle this problem, we present PropelWalker, a pair of calf-worn haptic devices for simulating the buoyancy and the resistant force when the human's lower limbs are interacting with different fluids and materials in VR. By using four ducted fans, two installed on each calf, the system can control the strength and the direction of the airflow in real time to provide different levels of force. Our technical evaluation shows that PropelWalker can generate vertical forces up to 27N in two directions (i.e., upward and downward) within 0.85 seconds. Furthermore, the system can stably maintain the generated force with minor turbulence. We further conducted three user-perception studies to understand the capability of PropelWalker to generate distinguishable force stimuli. First, we conducted the just-noticeable-difference (JND) experiments to investigate the threshold of the human perception of on-leg air-flow force feedback. Our second perception study showed that users could distinguish four PropelWalker-generated force levels for simulating different walking mediums (i.e., dry ground, water, mud, and sand), with an average accuracy of 94.2%. Lastly, our VR user study showed that PropelWalker could significantly improve the users' sense of presence in VR.

In this paper, we present a series of user studies to investigate the technique of gesture-amplitude-based walking-speed control for locomotion in place (LIP) in virtual reality (VR). Our 1st study suggested that compared to tapping and goose-stepping, the gesture of marching in place was significantly preferred by users across three different virtual walking speed (i.e., 1 ×, 3 ×, and 10 ×) while sitting and standing, and it yielded larger motion difference across the three speed levels. With the tracker data recorded in the 1st study, we trained a Support- Vector-Machine classification model for LIP speed control based on users' leg/foot gestures in marching in place. The overall accuracy for classifying three speed levels was above 90% for sitting and standing. With the classification model, we then compared the marching-in-place speed-control technique with the controller-based teleportation approach on a target-reaching task where users were sitting and standing. We found no significant difference between the two conditions in terms of target-reaching accuracy. More importantly, the technique of marching in place yielded significantly higher user ratings in terms of naturalness, realness, and engagement than the controller-based teleportation did.

We present ThermAirGlove (TAGlove), a pneumatic glove which provides thermal feedback for users, to support the haptic experience of grabbing objects of different temperatures and materials in virtual reality (VR). The system consists of a glove with five inflatable airbags on the fingers and the palm, two temperature chambers (one hot and one cold), and the closed-loop pneumatic thermal control system. Our technical experiments showed that the highest temperature-changing speed of TAGlove system was 2.75°C/s for cooling, and the pneumatic-control mechanism could generate the thermal cues of different materials (e.g., foam, glass, copper, etc.). The user-perception experiments showed that the TAGlove system could provide five distinct levels of thermal sensation (ranging from very cool to very warm). The user-perception experiments also showed that the TAGlove could support users’ material identification among foam, glass, and copper with the average accuracy of 87.2%, with no significant difference compared to perceiving the real physical objects. The user studies on VR experience showed that using TAGlove in immersive VR could significantly improve users’ experience of presence compared to the situations without any temperature or material simulation.