Smell and temperature are tightly intertwined in everyday experience, yet interactive systems often treat them as independent sensory channels. While thermohaptics and olfactory interfaces have been studied separately, it remains unclear how localized nasal thermal stimulation can modulate olfactory perception without changing chemical concentration. We introduce a nose-mounted thermohaptic interface that delivers precise, closed-loop heating and cooling stimuli to the nasal skin area (nasal sidewalls) to modulate olfactory perception. We evaluate our system in two user studies. In the first psychophysical study, we estimate perceptual thresholds (just-noticeable differences, JNDs) for thermally induced changes in perceived odor intensity across three odors, revealing intensity biasing and asymmetric sensitivity between cooling and heating (mean JND: 4.33°C for cooling; 6.15°C for heating). In the second study, we profile odor quality in a structured multidimensional descriptor space, revealing odor-dependent, dimension-specific semantic shifts rather than uniform changes. Together, these findings suggest that nasal thermal modulation can serve as a rapid, reversible, and non-chemical approach for shaping olfactory experience, expanding the design space for perceptually driven olfactory interfaces in multisensory Human-Computer Interaction (HCI).

Pet ownership creates profound human-animal bonds, making pet loss significant. However, compared to human loss, Human-Computer Interaction (HCI) has given pet loss less attention. Addressing this, we conducted an exploratory mixed-methods study. The study began with a large-scale survey (N=611) revealing critical challenges: a strong desire to preserve memories but limited support for continuing bonds. Built upon these findings, our participatory design sessions(N=10) co-designed Bondi, a tangible prototype supporting continuing bonds through multimodal and customizable interactions (e.g. touch, sound, and light), evoking pets’ unique sounds, tail movements, and lighting effects. We then conducted a three-week field-deployment study with four participants to evaluate how Bondi facilitated the maintenance of bonds with their deceased pets. Results showed that the customization and multimodality evoked vivid recollections, lowering the social barrier for grief sharing. Bondi fostered comforting and non-intrusive connections with pet memories. Furthermore, the study distilled design considerations for future pet bereavement support.

For VR interaction, the home environment with complicated spatial setup and dynamics may hinder the VR user experience. In particular, pets' movement may be more unpredictable. In this paper, we investigate the integration of real-world pet activities into immersive VR interaction. Our pilot study showed that the active pet movements, especially dogs, could negatively impact users' performance and experience in immersive VR. We proposed three different types of pet integration, namely semitransparent real-world portal, non-interactive object in VR, and interactive object in VR. We conducted the user study with 16 pet owners and their pets. The results showed that compared to the baseline condition without any pet-integration technique, the approach of integrating the pet as interactive objects in VR yielded significantly higher participant ratings in perceived realism, joy, multisensory engagement, and connection with their pets in VR.