Pith. sign in

REVIEW 3 major objections 4 minor 70 references

Talking Spell: A Wearable System Enabling Real-Time Anthropomorphic Voice Interaction with Everyday Objects

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Talking Spell claims that a wearable camera-and-touch kit can give any everyday object a persistent, AI-generated voice persona and deepen users' emotional attachment to it.

desk verdict A genuinely new wearable system with credible technical components, but the emotional-attachment claim rests on a 12-person, no-control study and needs far stronger evidence. read the letter →

arxiv 2509.02367 v1 pith:GB6QDZU2 submitted 2025-08-28 cs.HC

Xuetong Wang , Ching Christie Pang , Pan Hui This is my paper · ORCID
classification cs.HC
keywords AIcompanionshipanthropomorphicvoiceinteractionwearablehuman-objectobjectdetectionLLM-drivendialogueon-bodydevicespersonalization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Talking Spell sets out to prove that AI companionship does not require a new device: a wearable camera-and-touch kit lets users "cast a spell" on any everyday object they already own, giving it a name, persona, voice, and conversational memory. The system walks users through three stages—acquaintance (see and meet the object), familiarization (recognize it again and retrieve its persona), and bonding (talk to it and build a shared history). If the paper is right, the emotional bond people already feel toward a childhood toy, plant, or mug can be activated and deepened through ordinary speech, rather than transferred to an unfamiliar assistant. A 12-person user study reports high usability (SUS 79.09), companionship as the most-valued interaction intent, and participants describing objects as lifelong friends. The technical evaluation reports robust re-identification and near-real-time dialogue, supporting the feasibility of the pipeline.

What carries the argument

The load-bearing mechanism is the three-stage bond-building pipeline: Acquaintance (capture frames, segment the object, generate a persona JSON), Familiarization (train or reuse a small object detector to re-identify the object and retrieve its persona), and Bonding (wand-triggered voice chat with a language model, persona-driven text-to-speech, and a short-term memory buffer). Around this pipeline the paper places a "user-centric radiative network": the user is the centroid and each everyday object is a spoke, replacing the usual hub-and-spoke model in which the virtual assistant sits in a separate device. The three stages give the system a narrative structure for emotional connection, whil

What would settle it

A controlled comparison in which the same participant talks to the same personal object for equal time—once with Talking Spell and once with a plain voice-chat app or silent reflection—and fills the same attachment questionnaire would reveal whether the wearable pipeline itself changes reported attachment; if scores are equal, the central claim is unsupported.

Watch

Extended reading notes

Core claim

At its center, Talking Spell claims that the barrier to emotional AI companions is not the AI but the host: asking users to bond with an unfamiliar doll, speaker, or app is what makes companionship feel hollow. The paper's discovery is a working pipeline that attaches the AI to whatever object the user already cherishes. A scope (wearable camera) captures 100 frames of the object, a segmentation model extracts the object, a vision-language model invents a persona with name, age, personality, background story, and voice, and a compact detector is trained on the fly so the object is re-identified later. The wand (touch sensor plus vibration motor) lets the user hold-to-talk, and a language mod

Load-bearing premise

The paper's central claim rests on 12 participants, one 60-minute session, self-reported surveys, and no control condition; the load-bearing premise is that the emotional significance they reported was caused by Talking Spell rather than by already loving the object they brought in or by the novelty of the gadget.

Editorial extensions

If this is right

  • If the central claim holds, AI companions no longer need a dedicated form factor; any beloved object can become the interface, potentially increasing long-term engagement.
  • Because the system saves a separate persona and chat history per object, users can build a personal network of talking objects around themselves, each with a distinct voice and memory.
  • The reported real-time dialogue metrics (0.63 real-time factor, with long responses split for parallel TTS) suggest the interaction can keep pace with natural conversation rather than turn-taking with long delays.
  • Wearable form factors such as glasses, head chains, earrings, and bone-conduction headphones make the capability hands-free and unobtrusive.
  • The authors position the system for contexts—education, creative co-design, and elderly companionship—where emotional presence matters as much as task completion.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension is longitudinal: the paper's one-hour sessions leave open whether attachment grows, plateaus, or fades over weeks of daily use; a follow-up with repeated sessions would test this.
  • If the mechanism works on personal objects, the same pipeline could be pointed at public or shared objects (park benches, museum exhibits, tools), turning voice personas into a general ambient-computing layer rather than a personal-companion feature.
  • The on-the-fly training design suggests a testable floor: whether re-identification degrades when many similar objects are added, and whether the roughly 10-minute training time holds at scale.
  • The paper's own participants asked for proactive, multi-agent behavior; an implicit consequence is that the next version should coordinate multiple objects speaking to each other, transforming the radiative network from a hub into a social graph.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. Talking Spell is a wearable system that combines a camera-equipped 'scope' and a touch/vibration 'wand' with computer vision, vision-language models, speech-to-text, LLM, and TTS components to let users 'cast a spell' on everyday objects, giving them anthropomorphic personas and enabling spoken dialogue. The paper describes the hardware design, a three-stage interaction model (acquaintance, familiarization, bonding), a technical evaluation of object detection, persona generation, and dialogue speed, and a 12-participant user study that reports SUS, intent rankings, and qualitative feedback. The central claim is that Talking Spell 'fosters meaningful interactions and emotional significance with everyday objects' and creates engaging personalized experiences.

Significance. If validated, Talking Spell contributes a novel interaction paradigm: instead of bonding with a new AI device, users attach an AI persona to objects they already own and value. The technical pipeline is a clear strength: YOLOv11 re-identification reaches mAP 0.995 and real-world detection accuracy above 90% for 8 object classes, and the incremental training and dialogue response times are reported concretely. The wearable form-factor exploration (glasses, head chain, earrings, necklace, bone-conduction headphones) is useful design-space work. The paper is honest in calling the user study preliminary and in listing limitations in Section 7. However, the abstract and conclusion make causal claims about emotional effectiveness that the current evidence does not support; the user study lacks a control condition, uses self-selected cherished objects, relies on self-report, and is explicitly labeled preliminary. The appropriate scope of this paper is a system demonstration with promising usability and exploratory experiential findings, not a validated claim of induced emotional attachment.

major comments (3)
  1. [Abstract; Section 4; Section 7] The central claim that Talking Spell 'demonstrates its effectiveness in fostering meaningful interactions and emotional significance' is unsupported by the study as reported. Participants brought their own cherished items (Section 4.3.3), so quotes such as P10's 'it felt like I was talking to a lifelong friend' (Section 4.4.2) likely measure pre-existing attachment; there is no control condition, no baseline, and no manipulation check. The paper itself states the study is 'preliminary' (Section 7) and that 'empirical evidence is still needed' (Section 4.4.2). The abstract and conclusion should be reworded to describe the results as a preliminary usability and experience demonstration, or the study must be redesigned to control for pre-existing attachment and novelty.
  2. [Section 4.4.1, Table 4] The ranking data from 12 participants are used to draw conclusions such as 'anthropomorphism is generally more favored in recreational contexts than in practical ones.' This overstates what a frequency distribution from N=12 can support: no inferential test is reported, the four 'intents' were predefined by the authors (Sections 3.4 and 4.2) rather than discovered from user behavior, and the categories may steer participants' responses. The results should be framed as descriptive preferences within the specifically tested four-intent taxonomy, not as general evidence about anthropomorphism preferences.
  3. [Section 3.5.2 (Persona Creation)] The claim that QWEN-VL 'outperforms' GPT-3.5 and Grok-3 in persona generation rests on ratings from three participants with no inter-rater reliability, no statistical test, and a small number of objects. Given the large standard deviations (e.g., Grok-3 SD = 1.6854), the mean differences may not be reliable. Please report individual ratings, add an appropriate test or confidence interval, and soften the conclusion to 'preliminary evidence' consistent with the rest of the paper.
minor comments (4)
  1. [Section 5] The 'Open Source Application' section says circuit designs are 'not available now due to the review anonymous consideration,' which contradicts the earlier repository link and the stated goal of enabling replication. If the anonymous-review version requires a placeholder, say so explicitly, or provide the actual open-source materials before publication.
  2. [Section 3.3.3] 'Azure ChatGPT Turbo-3.5' is ambiguous; please use the exact model name (e.g., Azure OpenAI GPT-3.5-Turbo) and version.
  3. [Section 6.5] Typo: 'wearabke systems' should be 'wearable systems.'
  4. [Section 3.2.1] The sentence 'we selected three sizes of the OV2640 camera module: 2.1cm, 7.5cm, and 20cm' likely refers to cable lengths or form-factor variants, not the sensor die; please clarify the dimension being described.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: central claims are empirical and externally evaluated; only minor instrument-design self-reference.

full rationale

Talking Spell is a systems paper whose claims are empirical rather than derived: it builds hardware, connects off-the-shelf vision/LLM/TTS components, and evaluates them with a technical benchmark and a 12-person user study. No equation-level derivation chain is present, so no result is equivalent to an input by construction. The detection pipeline is measured against real frames (Tables 2-3); even though the test objects belong to trained categories, the reported accuracies are empirical outcomes that could have failed and are not statistically forced by the training procedure. Persona generation is compared across three independent models with blinded participant ratings. The user study is external, and the paper self-limits its conclusions ('preliminary evidence', Section 7; 'empirical evidence is still needed', Section 4.4.2), which mitigates concern that the emotional-attachment claim is presented as a forced deduction. The only mild self-referential element is that the authors designed both the four-intent taxonomy and the predefined objects used in the ranking task (Section 4.2), so the instrument may nudge the categories it reports; however, this is a framing/validity issue, not a logical circle, because participant rankings could have contradicted the taxonomy and the paper does not invoke any self-citation or uniqueness theorem. No load-bearing self-citations, no fitted parameter renamed as prediction, and no ansatz smuggled in via citation were found. Score 1 reflects a minor design confound with no circularity.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

This is a systems paper, not a derivation; the ledger records hand-chosen design parameters and domain assumptions that the effectiveness claim rests on. No new physical entities are postulated.

free parameters (5)
  • Detection confidence threshold = >0.75
    Used in Table 3 to categorize successful detections; a lower threshold would change reported accuracy.
  • Persona attribute schema and prompt template
    Hand-authored prompt to QWEN-VL determines the generated name, gender, age, personality, story, and voice; this shapes the emotional bonding experience.
  • Voice inventory (7 voices) = 0_old_woman, 1_young_woman, 2_kid_woman, 3_old_man, 4_young_man, 5_kid_man, 6_no_gender
    Hand-selected TTS timbre set; participants criticized its limited variety and repetition, so the emotional experience depends on this choice.
  • Memory buffer length = 10 records, LIFO after 5 cycles
    Arbitrarily chosen in Section 3.3.3; controls how much conversational history the persona can use.
  • Training epochs and early stopping patience = 100 epochs, patience=25
    Chosen training hyperparameters in Section 3.3.2; affect the re-identification model that the Familiarization stage depends on.
assumptions (6)
  • domain assumption Anthropomorphism increases emotional engagement with objects
    Invoked throughout Section 2.4 and used to justify the design; based on prior literature but not re-established here.
  • domain assumption Voice interaction alone is sufficient for meaningful anthropomorphic connection
    The system provides no visual anthropomorphic features on the object itself; the persona exists only in voice and dialogue. Section 4.4.2 reports voice importance, but as a post-hoc finding.
  • domain assumption The acquaintance-familiarization-bonding framework maps to real attachment dynamics
    The three-stage model in Section 3.3 is presented as a design analogy, not validated against psychological attachment theory.
  • domain assumption Self-reported SUS, NPS, rankings, and interviews measure emotional significance
    Section 4.4 uses these instruments to conclude effectiveness; no physiological or behavioral measures are used.
  • domain assumption Object re-identification persists beyond the test session
    Table 3 evaluates only 200 consecutive frames; the cross-session Familiarization claim is not tested.
  • domain assumption Cloud AI APIs (QWEN-VL, ChatGPT, Faster Whisper, Coqui TTS) behave reliably as black boxes
    The system's correctness depends on external services whose behavior and availability are outside the paper's control.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Talking Spell: A Wearable System Enabling Real-Time Anthropomorphic Voice Interaction with Everyday Objects." pith.science (2026). https://pith.science/paper/GB6QDZU2

@misc{pith2026250902367,
  author       = {Pith},
  title        = {Pith review of: Talking Spell: A Wearable System Enabling Real-Time Anthropomorphic Voice Interaction with Everyday Objects},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GB6QDZU2}},
  note         = {Machine review of arXiv:2509.02367}
}
read the original abstract

Virtual assistants (VAs) have become ubiquitous in daily life, integrated into smartphones and smart devices, sparking interest in AI companions that enhance user experiences and foster emotional connections. However, existing companions are often embedded in specific objects-such as glasses, home assistants, or dolls-requiring users to form emotional bonds with unfamiliar items, which can lead to reduced engagement and feelings of detachment. To address this, we introduce Talking Spell, a wearable system that empowers users to imbue any everyday object with speech and anthropomorphic personas through a user-centric radiative network. Leveraging advanced computer vision (e.g., YOLOv11 for object detection), large vision-language models (e.g., QWEN-VL for persona generation), speech-to-text and text-to-speech technologies, Talking Spell guides users through three stages of emotional connection: acquaintance, familiarization, and bonding. We validated our system through a user study involving 12 participants, utilizing Talking Spell to explore four interaction intentions: entertainment, companionship, utility, and creativity. The results demonstrate its effectiveness in fostering meaningful interactions and emotional significance with everyday objects. Our findings indicate that Talking Spell creates engaging and personalized experiences, as demonstrated through various devices, ranging from accessories to essential wearables.

Figures

Figures reproduced from arXiv: 2509.02367 by the authors.

Figure 1
Figure 1. Talking Spell consists the scope (camera for detect [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The exploded view diagram of the scope’s compo [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Wand Design. Top: Circuit schematic diagram. Bot [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (16 more)
Figure 6
Figure 6. Figure 6: Usage flow of the bonding stage. signal is sent back to the wand, simultaneously activating the vi￾bration motor to provide haptic feedback. The recording ceases once the user releases the touch sensor, and the audio is saved as an MP3 file. This MP3 file is then proce…
Figure 5
Figure 5. Figure 5: Create anthropomorphic persona using large visual [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 7
Figure 7. Figure 7: Flowchart of the bonding stage. Oval:start/end; [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 10
Figure 10. Figure 10: Stages in Entertainment with Talking Spell: from [PITH_FULL_IMAGE:figures/full_fig_p007_10.png]
Figure 8
Figure 8. Figure 8: Talking Spell encompasses four primary interac [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 12
Figure 12. Figure 12: Stages in Companion with Talking Spell: from [PITH_FULL_IMAGE:figures/full_fig_p008_12.png]
Figure 13
Figure 13. Figure 13: Stages in Utility with Talk Spell: from a decoration [PITH_FULL_IMAGE:figures/full_fig_p009_13.png]
Figure 15
Figure 15. Figure 15: Example Environments for Evaluation Experi [PITH_FULL_IMAGE:figures/full_fig_p009_15.png]
Figure 16
Figure 16. Figure 16: Talking Spell emphasizes the accessibility and ver [PITH_FULL_IMAGE:figures/full_fig_p010_16.png]
Figure 18
Figure 18. Figure 18: Sketches by the participants with the help from [PITH_FULL_IMAGE:figures/full_fig_p012_18.png]
Figure 17
Figure 17. Figure 17: Participant (N=12) feedback on the seven assess [PITH_FULL_IMAGE:figures/full_fig_p012_17.png]
Figure 19
Figure 19. Figure 19: A bone conduction headphone that serves as the [PITH_FULL_IMAGE:figures/full_fig_p013_19.png]
Figure 22
Figure 22. Figure 22: A Talking Spell head chain and glasses, allowing [PITH_FULL_IMAGE:figures/full_fig_p014_22.png]
Figure 20
Figure 20. Figure 20: A bone conduction headphone as an example of [PITH_FULL_IMAGE:figures/full_fig_p014_20.png]
Figure 23
Figure 23. Figure 23: A Talking Spell wand in form of charm, earrings, [PITH_FULL_IMAGE:figures/full_fig_p014_23.png]
Figure 21
Figure 21. Figure 21: A Talking Spell scope in form of head chain and [PITH_FULL_IMAGE:figures/full_fig_p014_21.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

70 extracted references · 55 canonical work pages

  1. [1]

    Alkatheiri

    Mohammed S. Alkatheiri. 2022. Artificial intelligence assisted improved human- computer interactions for computer systems. Computers and Electrical Engineer- ing 101 (2022), 107950. doi:10.1016/j.compeleceng.2022.107950

    work page arXiv 2022

  2. [2]

    Tsai, and Frank Bentley

    Tawfiq Ammari, Jofish Kaye, Janice Y. Tsai, and Frank Bentley. 2019. Music, Search, and IoT: How People (Really) Use Voice Assistants. ACM Trans. Comput.- Hum. Interact. 26, 3, Article 17 (April 2019), 28 pages. doi:10.1145/3311956

    doi:10.1145/3311956 2019

  3. [3]

    Michele Farisco and Arleen Salles, Kathinka Evers. 2020. Anthropomorphism in AI. AJOB Neuroscience 11, 2 (2020), 88–95. doi:10.1080/21507740.2020.1740350 arXiv:https://doi.org/10.1080/21507740.2020.1740350 PMID: 32228388

    arXiv 2020

  4. [4]

    Jinze Bai, Shuai Bai, Shusheng Yang, Shijie Wang, Sinan Tan, Peng Wang, Jun- yang Lin, Chang Zhou, and Jingren Zhou. 2023. Qwen-VL: A Versatile Vision- Language Model for Understanding, Localization, Text Reading, and Beyond. arXiv:2308.12966 [cs.CV] https://arxiv.org/abs/2308.12966

    arXiv 2023

  5. [5]

    Reginald Boersma, P Marijn Poortvliet, and Bart Gremmen. 2019. Naming is framing: the effects of a technological name on the interpretation of a technology. Journal of Science Communication 18, 6 (2019), A04

    work page 2019

  6. [6]

    John Bowlby. 1969. Attachment and loss. Number 79. Random House

    work page 1969

  7. [7]

    John Brooke et al. 1996. SUS-A quick and dirty usability scale.Usability evaluation in industry 189, 194 (1996), 4–7

    work page 1996

  8. [8]

    Julie Carpenter. 2013. Just doesn’t look right: Exploring the impact of humanoid robot integration into explosive ordnance disposal teams. InHandbook of research on technoself: Identity in a technological society . IGI Global, 609–636

    work page 2013

Show all 70 references
  1. [9]

    Daniela Castillo, Ana Isabel Canhoto, and Emanuel Said and. 2021. The dark side of AI-powered service interactions: exploring the process of co-destruction from the customer perspective. The Service Industries Journal 41, 13-14 (2021), 900–925. doi:10.1080/02642069.2020.178799...

    2021

  2. [10]

    Yu-Wei Chao, Yunfan Liu, Xieyang Liu, Huayi Zeng, and Jia Deng. 2018. Learning to detect human-object interactions. In 2018 ieee winter conference on applications of computer vision (wacv). IEEE, 381–389

    2018

  3. [11]

    Wen-Wei Cheng and Liwei Chan. 2024. BodyTouch: Investigating Eye-Free, On- Body and Near-Body Touch Interactions with HMDs. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 7, 4, Article 152 (Jan. 2024), 22 pages. doi:10.1145/ 3631426

    2024

  4. [12]

    Terrence Chong, Ting Yu, Debbie Isobel Keeling, and Ko de Ruyter. 2021. AI- chatbots on the services frontline addressing the challenges and opportunities of agency. Journal of Retailing and Consumer Services 63 (2021), 102735

    2021

  5. [13]

    What can i help you with?

    Benjamin R Cowan, Nadia Pantidi, David Coyle, Kellie Morrissey, Peter Clarke, Sara Al-Shehri, David Earley, and Natasha Bandeira. 2017. " What can i help you with?" infrequent users’ experiences of intelligent personal assistants. In Proceedings of the 19th international confe...

    2017

  6. [14]

    Alba Cercas Curry and Amanda Cercas Curry. 2023. Computer says “no”: The case against empathetic conversational AI. In Findings of the Association for Computational Linguistics: ACL 2023. 8123–8130

    2023

  7. [15]

    Shiyu Duan, Ziyi Wang, Shixiao Wang, Mengmeng Chen, and Runsheng Zhang

  8. [16]

    Fotis Efthymiou, Alex Mari, Ertugrul Uysal, and Jeffrey Brooks. 2025. No Hard Feelings: The Protective Power of AI Empathy During Service Interaction Failures. (2025)

    2025

  9. [17]

    Nicholas Epley, Adam Waytz, and John T Cacioppo. 2007. On seeing human: a three-factor theory of anthropomorphism. Psychological review 114, 4 (2007), 864

    2007

  10. [18]

    A Shaji George, AS Hovan George, T Baskar, and Digvijay Pandey. 2023. The allure of artificial intimacy: Examining the appeal and ethics of using generative AI for simulated relationships. Partners Universal International Innovation Journal 1, 6 (2023), 132–147

    2023

  11. [19]

    Marlène Gilles and Elisabetta Bevacqua. 2022. A review of virtual assistants’ characteristics: Recommendations for designing an optimal human–machine cooperation. Journal of Computing and Information Science in Engineering 22, 5 (2022), 050904

    2022

  12. [20]

    Georgia Gkioxari, Ross Girshick, Piotr Dollár, and Kaiming He. 2018. Detecting and recognizing human-object interactions. In Proceedings of the IEEE conference on computer vision and pattern recognition . 8359–8367

    2018

  13. [21]

    Ella Glikson and Anita Williams Woolley. 2020. Human trust in artificial intelli- gence: Review of empirical research. Academy of management annals 14, 2 (2020), 627–660

    2020

  14. [22]

    Frank Hegel, Soren Krach, Tilo Kircher, Britta Wrede, and Gerhard Sagerer

  15. [23]

    Pauline J Horne and C Fergus Lowe. 1996. On the origins of naming and other symbolic behavior. Journal of the Experimental Analysis of behavior 65, 1 (1996), 185–241

    1996

  16. [24]

    Xiyun Hu, Dizhi Ma, Fengming He, Zhengzhe Zhu, Shao-Kang Hsia, Chenfei Zhu, Ziyi Liu, and Karthik Ramani. 2025. GesPrompt: Leveraging Co-Speech Gestures to Augment LLM-Based Interaction in Virtual Reality. In Proceedings of the 2025 ACM Designing Interactive Systems Conference...

    2025

  17. [25]

    Złotowski Jakub, Diane Proudfoot, Yogeeswaran Kumar, and Bartneck Christoph

  18. [26]

    Mohammad Hossein Jarrahi. 2018. Artificial intelligence and the future of work: Human-AI symbiosis in organizational decision making. Business horizons 61, 4 (2018), 577–586

    2018

  19. [27]

    Hannah Rose Kirk, Iason Gabriel, Chris Summerfield, Bertie Vidgen, and Scott A Hale. 2025. Why human-AI relationships need socioaffective alignment. arXiv preprint arXiv:2502.02528 (2025)

    2025 arXiv

  20. [28]

    Rinaldo Kühne and Jochen Peter. 2022. Anthropomorphism in hu- man–robot interactions: a multidimensional conceptualization. Communi- cation Theory 33, 1 (Oct. 2022), 42–52. doi:10.1093/ct/qtac020 tex.eprint: https://academic.oup.com/ct/article-pdf/33/1/42/48958537/qtac020.pdf

    2022 doi

  21. [29]

    Rodriguez, and Jon E

    Jaewook Lee, Jun Wang, Elizabeth Brown, Liam Chu, Sebastian S. Rodriguez, and Jon E. Froehlich. 2024. GazePointAR: A Context-Aware Multimodal Voice Assis- tant for Pronoun Disambiguation in Wearable Augmented Reality. InProceedings of the 2024 CHI Conference on Human Factors i...

    2024

  22. [30]

    You Shouldn’t Need to Share Your Data

    Anna Lenhart, Sunyup Park, Michael Zimmer, and Jessica Vitak. 2023. "You Shouldn’t Need to Share Your Data": Perceived Privacy Risks and Mitigation Strategies Among Privacy-Conscious Smart Home Power Users.Proc. ACM Hum.- Comput. Interact. 7, CSCW2, Article 247 (Oct. 2023), 34...

    2023 doi

  23. [31]

    Peng Liu, Qingqing Fei, Jinting Liu, and Jianqiang Wang. 2021. Naming is framing: The framing effect of technology name on public attitude toward automated vehicles. Public Understanding of Science 30, 6 (2021), 691–707

    2021

  24. [32]

    Alexander Maedche, Christine Legner, Alexander Benlian, Benedikt Berger, Hen- ner Gimpel, Thomas Hess, Oliver Hinz, Stefan Morana, and Matthias Söllner

  25. [33]

    Whaf s in a Name?

    Sally Mcconnell-Ginet. 2005. "Whaf s in a Name?" Social Labeling and Gender Practices. The Handbook of Language and Gender (2005), 69

    2005

  26. [34]

    Mario Mikulincer and P Shaver. 2005. Mental representations and attachment security. Interpersonal cognition (2005), 233–266

    2005

  27. [35]

    Julia Mossbridge. 2024. Shifting the Human-AI Relationship: Toward a Dynamic Relational Learning-Partner Model. arXiv preprint arXiv:2410.11864 (2024)

    2024 arXiv

  28. [36]

    Prudhvi Naayini, Praveen Kumar Myakala, Chiranjeevi Bura, Anil Kumar Jon- nalagadda, and Srikanth Kamatala. 2025. AI-Powered Assistive Technologies for Visual Impairment. arXiv preprint arXiv:2503.15494 (2025)

    2025 arXiv

  29. [37]

    Donald A Norman. 2010. Natural user interfaces are not natural. interactions 17, 3 (2010), 6–10

    2010

  30. [38]

    Kohei Ogawa and Tetsuo Ono. 2008. ITACO: Constructing an emotional re- lationship between human and robot. In RO-MAN 2008 - The 17th IEEE Inter- national Symposium on Robot and Human Interactive Communication . 35–40. doi:10.1109/ROMAN.2008.4600640

    2008

  31. [39]

    Kenton O’hara, Richard Harper, Helena Mentis, Abigail Sellen, and Alex Taylor

  32. [40]

    Hirotaka Osawa, Kentaro Ishii, Toshihiro Osumi, Ren Ohmura, and Michita Imai. 2009. Anthropomorphization of a space with implemented human-like features. In ACM SIGGRAPH 2009 Emerging Technologies(New Orleans, Louisiana) (SIGGRAPH ’09). Association for Computing Machinery, New...

    2009

  33. [41]

    Hirotaka Osawa, Yuji Matsuda, Ren Ohmura, and Michita Imai. 2012. Embodiment of an agent by anthropomorphization of a common object. Web Intelligence and Agent Systems 10, 3 (2012), 345–358

    2012

  34. [42]

    Hirotaka Osawa, Jun Mukai, and Michita Imai. 2006. Anthropomorphization of an object by displaying robot. In ROMAN 2006-The 15th IEEE International Symposium on Robot and Human Interactive Communication . IEEE, 763–768

    2006

  35. [43]

    Diane Proudfoot. 2011. Anthropomorphism and AI: Turing’s much misunder- stood imitation game. Artificial Intelligence 175, 5-6 (2011), 950–957

    2011

  36. [44]

    Alexa is my new BFF

    Amanda Purington, Jessie G Taft, Shruti Sannon, Natalya N Bazarova, and Samuel Hardman Taylor. 2017. " Alexa is my new BFF" social roles, user satis- faction, and personification of the Amazon Echo. In Proceedings of the 2017 CHI conference extended abstracts on human factors ...

    2017

  37. [45]

    Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman Rädle, Chloe Rolland, Laura Gustafson, et al. 2024. Sam 2: Segment anything in images and videos. arXiv preprint arXiv:2408.00714 (2024)

    2024 arXiv

  38. [46]

    Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. 2016. You only look once: Unified, real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition . 779–788

    2016

  39. [47]

    Byron Reeves and Clifford Nass. 1996. The media equation: How people treat computers, television, and new media like real people. Cambridge, UK 10, 10 (1996), 19–36

    1996

  40. [48]

    Karen S Rook. 1987. Social support versus companionship: effects on life stress, loneliness, and evaluations by others. Journal of personality and social psychology 52, 6 (1987), 1132

    1987

  41. [49]

    Mahnaz Roshanaei, Rezvaneh Rezapour, and Magy Seif El-Nasr. 2024. Talk, Listen, Connect: Navigating Empathy in Human-AI Interactions. arXiv preprint arXiv:2409.15550 (2024)

    2024

  42. [50]

    Mehdi Rostami and Shokouh Navabinejad. 2023. Artificial empathy: User experi- ences with emotionally intelligent chatbots. AI and Tech in Behavioral and Social Sciences 1, 3 (2023), 19–27

    2023

  43. [51]

    Lin, Adam S

    Ashish Sharma, Inna W. Lin, Adam S. Miner, David C. Atkins, and Tim Althoff

  44. [52]

    Dara Sorkin, Karen S Rook, and John L Lu. 2002. Loneliness, lack of emotional support, lack of companionship, and the likelihood of having a heart condition in an elderly sample. Annals of Behavioral Medicine 24, 4 (2002), 290–298

    2002

  45. [53]

    Jennifer L Stoner, Barbara Loken, and Ashley Stadler Blank. 2018. The name game: How naming products increases psychological ownership and subsequent consumer evaluations. Journal of Consumer Psychology 28, 1 (2018), 130–137

    2018

  46. [54]

    Megan Su and Yuwei Bao. 2024. User Modeling Challenges in Interactive AI Assis- tant Systems. In The Twelfth International Conference on Learning Representations: Tiny Papers Track

    2024

  47. [55]

    Ravi Tejwani. 2020. Migratable AI. Ph. D. Dissertation. Massachusetts Institute of Technology

    2020

  48. [56]

    Ravi Tejwani, Felipe Moreno, Sooyeon Jeong, Hae Won Park, and Cynthia Breazeal. 2020. Migratable AI: Effect of identity and information migration on users’ perception of conversational AI agents. In 2020 29th IEEE International Conference on Robot and Human Interactive Communi...

    2020

  49. [57]

    TINA Ten Bruggencate, Katrien G Luijkx, and Janienke Sturm. 2018. Social needs of older people: A systematic literature review. Ageing & Society 38, 9 (2018), 1745–1770

    2018

  50. [58]

    Radu-Daniel Vatavu. 2025. Non-Natural Interaction Design. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI ’25) . As- sociation for Computing Machinery, New York, NY, USA, Article 435, 16 pages. doi:10.1145/3706598.3713459

    2025

  51. [59]

    Echo Wen Wan and Rocky Peng Chen. 2021. Anthropomorphism and object attachment. Current Opinion in Psychology 39 (2021), 88–93. doi:10.1016/j.copsyc. 2020.08.009 Object Attachment

    2021 doi

  52. [60]

    Cara Wrigley, Sean Peel, Kimmi Keum Hee Ko, and Karla Straker. 2021. Patient names for mechanical circulatory support devices: Developing emotional insights. Heart & Lung 50, 6 (2021), 953–967

    2021

  53. [61]

    Heetae Yang and Hwansoo Lee. 2019. Understanding user behavior of virtual personal assistant devices. Information Systems and e-Business Management 17 (2019), 65–87

    2019

  54. [62]

    Ziqi Yang, Xuhai Xu, Bingsheng Yao, Ethan Rogers, Shao Zhang, Stephen Intille, Nawar Shara, Guodong Gordon Gao, and Dakuo Wang. 2024. Talk2Care: An LLM-based Voice Assistant for Communication between Healthcare Providers and Older Adults. Proc. ACM Interact. Mob. Wearable Ubiq...

    2024 doi

  55. [63]

    Yibing Zhan, Jun Yu, Ting Yu, and Dacheng Tao. 2019. On exploring undetermined relationships for visual relationship detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 5128–5137

    2019

  56. [64]

    Shuning Zhang and Shixuan Li. 2025. The Real Her? Exploring Whether Young Adults Accept Human-AI Love. arXiv preprint arXiv:2503.03067 (2025)

    2025 arXiv

  57. [2008]

    In RO- MAN 2008-the 17th IEEE international symposium on robot and human interactive communication

    Understanding social robots: A user study on anthropomorphism. In RO- MAN 2008-the 17th IEEE international symposium on robot and human interactive communication. IEEE, 574–579

    2008

  58. [2013]

    interaction

    On the naturalness of touchless: Putting the “interaction” back into NUI. ACM Trans. Comput.-Hum. Interact. 20, 1, Article 5 (April 2013), 25 pages. doi:10. 1145/2442106.2442111

    2013

  59. [2015]

    International Journal of Social Robotics 7, 3 (2015), 347–360

    Anthropomorphism: opportunities and challenges in human–robot inter- action. International Journal of Social Robotics 7, 3 (2015), 347–360

    2015

  60. [2019]

    Business & Information Systems Engineering 61 (2019), 535–544

    AI-based digital assistants: Opportunities, threats, and research perspec- tives. Business & Information Systems Engineering 61 (2019), 535–544

    2019

  61. [2023]

    Nature Machine Intelligence 5, 1 (Jan

    Human–AI collaboration enables more empathic conversations in text- based peer-to-peer mental health support. Nature Machine Intelligence 5, 1 (Jan. 2023), 46–57. doi:10.1038/s42256-022-00593-2

    2023 doi

  62. [2024]

    In 2024 5th International Symposium on Computer Engineer- ing and Intelligent Communications (ISCEIC)

    Emotion-aware interaction design in intelligent user interface using multi- modal deep learning. In 2024 5th International Symposium on Computer Engineer- ing and Intelligent Communications (ISCEIC) . IEEE, 110–114

    2024

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.