Scensory: Multisensory Integration & AI
- Scensory is a multidisciplinary concept that integrates olfaction, auditory, tactile, and visual modalities to enable enhanced multisensory perception and interaction.
- It underpins diverse applications, from robot-enabled fungal detection and VR multisensory design to theoretical frameworks for cross-modal communication.
- The research advances multisensory AI by merging innovative sensing technologies, rigorous scientific quantification, and synergistic integration across domains.
Searching arXiv for papers using or defining “Scensory”. I’ll check for arXiv entries and adjacent multisensory uses of the term. Scensory is a recent label used in several related ways across multisensory research. In one usage, it is the proper name of a robot-enabled olfactory system for real-time fungal identification and spatial mapping from volatile organic compound dynamics (Liu et al., 11 Sep 2025). In another, it denotes a multisensory design strategy that augments visually constrained virtual reality with sound and scent (Bak et al., 14 Sep 2025). In broader theoretical work, it appears as a shorthand for “Sensing, Science, and Synergy” and, separately, as a playful term for sensory perception understood as a universal communicative substrate across art, human–computer interaction, and artificial intelligence (Liang, 8 Jan 2026, Kang et al., 2024). Across these usages, the common theme is the treatment of smell, sound, touch, taste, and other non-visual channels as structured modalities for inference, interaction, or communication rather than as peripheral embellishments.
1. Term, scope, and major usages
The term does not denote a single standardized framework. Instead, recent literature uses it to name a specific sensing platform, to describe a particular multisensory interface strategy, and to articulate a general research program for multisensory intelligence.
| Usage | Description | Representative source |
|---|---|---|
| Proper system name | Robot-enabled olfactory monitoring for fungal species detection and localization | (Liu et al., 11 Sep 2025) |
| Interface strategy | Vision augmented with auditory and olfactory cues in VR portals | (Bak et al., 14 Sep 2025) |
| Theoretical shorthand | “Sensing, Science, and Synergy” for multisensory AI | (Liang, 8 Jan 2026) |
| Playful conceptual term | Sensory perception as a universal language across disciplines | (Kang et al., 2024) |
| Cross-sensory inference label | Prediction of taste, smell, texture, and sound from food images | (Ishraq et al., 15 Apr 2026) |
This suggests that “Scensory” functions less as a settled taxonomy than as a family of closely related concepts centered on multisensory representation, cross-modal correspondence, and the computational use of non-visual signals. The variability of usage is itself informative: some authors foreground olfaction, others multimodal interaction, and others a general expansion of AI beyond text, vision, and audio.
2. Conceptual formulations
A major theoretical formulation treats sensory perception as a universal, non-verbal medium linking otherwise separate disciplinary languages. Kang et al. argue that practitioners across art, science, and engineering face communication barriers, and adapt Seleskovitch’s Interpretive Theory of Translation into the mirrored relation “Source Language → Sense → Target Language” and “(Art language) → Scensory → (Engineering language)” (Kang et al., 2024). In this account, vision, hearing, touch, taste, and scent constitute a shared substrate through which meaning can be translated across media art, HCI, communication, translation studies, and AI. Their examples, ranging from Kandinsky and Xenakis to David Bowen’s Telepresent Water and Ned Kahn’s Wind Veil, frame sensory feedback not as ornament but as the syntax by which meaning is encoded and decoded.
A second formulation is explicitly programmatic. The multisensory AI vision paper defines “Scensory” as shorthand for the next stage of intelligence, one that fuses Sensing, Science, and Synergy into end-to-end systems (Liang, 8 Jan 2026). In this view, AI should extend beyond camera, microphone, and keyboard to “language, vision, sound, touch, taste, smell, physiology, environment, and social context.” The paper organizes the field around three problems. “Sensing” concerns novel hardware and representation pipelines for heterogeneous signals such as gas-sensor arrays, tactile traces, thermal-infrared scans, and physiological streams. “Science” concerns the quantification of heterogeneity, redundancy, uniqueness, and synergy across modalities, including information-theoretic decompositions such as
“Synergy” concerns integration, alignment, reasoning, generation, generalization, and situated co-experience between humans and AI (Liang, 8 Jan 2026).
Taken together, these theoretical uses position Scensory as both an epistemic claim and a design orientation. The epistemic claim is that perception itself can mediate between domains; the design orientation is that systems