Pith. sign in

REVIEW 3 major objections 7 minor 80 references

RemiAssist: A Therapist-Supporting System for Photo-Based Reminiscence Therapy in Dementia Care

T0 review · 3 major / 7 minor · reviewed 2026-07-31 · grok-4.5

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

Pith's one-line read AI that backs therapists—not replaces them—cuts PRT planning time and lengthens reminiscence talks with people with dementia.

desk verdict Solid therapist-in-the-loop PRT system with a real field deployment; the 44%/54% headlines are real associations but mostly measure scaffolding vs a bare manual baseline, which the authors already admit. read the letter →

arxiv 2607.24536 v2 pith:B4QWHT62 submitted 2026-07-27 cs.HC

Shuchang Xu , Minglong Tang , Junyan Mao , Xiaofu Jin , Wazeer Zulfikar , Yasith Samaradivakara , Jiayi Zhou , Huamin Qu
show 2 more authors
Yuling Sun Pattie Maes
This is my paper
classification cs.HC
keywords ReminiscenceTherapyDementiaHumanMemoryAI-MediatedCommunicationPhotoCollectionTherapist-in-the-loopContext-awaresystemsAccessibility
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

Photo-based reminiscence therapy helps people with dementia recall and share life stories, but therapists must plan personalized sessions and steer fragile live conversations. Most AI tools try to talk to patients directly and sidestep that clinical work. RemiAssist instead keeps the therapist in control: a Memory Graph turns a personal photo collection into theme-centered session plans, and a Context-Aware Guiding Strategy offers glanceable prompts, related photos, and gentle redirects when talk drifts or hits distress. In a field study with eight therapist–patient pairs, therapists planned faster, conversations ran longer, mental load fell, and therapists still felt they owned the session. The paper’s point is practical: AI that organizes life themes and supports real-time judgment can make personalized reminiscence therapy more reachable without removing the therapist from the loop.

What carries the argument

Memory Graph: a three-level hierarchy (Themes → Photos → contextual details of time, place, people, objects, activities) built from a photo collection so shared contexts link events and Leiden communities become reminiscence themes that seed editable session plans. Context-Aware Guiding Strategy: real-time guiding questions, relevant photos, and sensitivity prompts scored from live talk plus the graph and a family-supplied distress list, shown for at-a-glance therapist use.

What would settle it

Repeat the within-subject comparison with an equal-information, non-context-aware baseline (same plan detail and prompt volume, no live context matching) in a larger longitudinal sample; if planning time and conversation length no longer differ, the central attribution fails.

Watch

Extended reading notes

Core claim

Therapist-in-the-loop AI for photo-based reminiscence therapy—built around a hierarchical Memory Graph for theme-centered planning and a Context-Aware Guiding Strategy for live facilitation—was associated in a within-subject field study with about 44% faster session planning, about 54% longer reminiscence conversations, lower therapist mental demand, and usable support for sensitive moments, while therapists reported that suggestions remained optional rather than directive.

Load-bearing premise

The measured gains come from the graph and context-aware design themselves, not mainly from giving therapists more pre-written structure and AI content than the manual baseline, or from novelty in a small short study.

Editorial extensions

If this is right

  • AI for dementia reminiscence can target therapist workflow—draft theme plans before sessions and glanceable live cues during them—rather than full patient–AI replacement.
  • Organizing personal photos as a multi-level graph of shared people, places, and activities can make theme-centered PRT planning faster without forcing therapists to accept unedited AI plans.
  • Real-time suggestions that therapists can ignore preserve professional agency while still lengthening engagement and easing redirects off-topic or into distress.
  • Session summaries of recalled events and emotions can feed longer-term memory profiles for safer future topic choice.
  • Sensitivity support should stay tunable to each therapist’s judgment about when listening through painful emotion is better than immediate topic change.

Reading between the lines

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

  • The same Memory Graph could double as a family-facing life-story map if privacy and accuracy gates stay therapist- or family-mediated, not patient-facing for unverified AI labels.
  • Detecting proactive-vs-reactive and focused-vs-drifting styles mid-session (as the discussion sketches) would let suggestion strength ramp up or down automatically.
  • Equal-information ablation is the cleanest next experiment; without it, product claims risk over-crediting ‘context awareness’ versus simply more scaffolding.
  • Local or on-device models for the graph and live scoring would address the privacy tension the paper flags for long-term personal memory profiles.
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 / 7 minor

Summary. The paper presents RemiAssist, a web-based system that supports therapists (rather than replacing them) in delivering photo-based reminiscence therapy to people with early-stage dementia. It contributes (1) a Memory Graph — a hierarchical Themes→Photos→Context representation automatically constructed from a PwD's photo collection using face clustering, VLM-based context extraction, and Leiden community detection — used to auto-generate editable, theme-centered session plans; and (2) a Context-Aware Guiding Strategy that streams guiding questions, relevant photos, and sensitivity prompts during live sessions, ranked by a weighted scoring function with a large distress-avoidance penalty. Design considerations derive from a formative interview study with five therapists. A within-subject field study with eight therapist–PwD dyads (seven completing facilitation) compares RemiAssist against a baseline mimicking manual practice; the authors report a 44% reduction in planning time, a 54% increase in conversation duration, lower NASA-TLX mental demand, high plan-element accuracy (90–93%), and preserved therapist agency. The authors explicitly frame statistics as indicative trends given N=8 and uncorrected multiple comparisons.

Significance. If the results hold, the work matters on two fronts. Clinically adjacent: RT delivery is labor-intensive and therapist-scarce, and a system that demonstrably reduces preparation burden while preserving therapist agency addresses a real bottleneck; the paper is unusually careful about the ethical texture of the setting (family-vetted photo curation, distress-topic lists, therapist-controlled termination, no direct PwD self-report). Scientifically: the Memory Graph is a concrete, reusable representation for organizing personal photo archives around themes rather than timelines, and the real-time guiding strategy is a worked example of glanceable, suggestive-rather-than-directive AI in live conversation. Strengths worth naming: the design is grounded in a formative study with practitioners; the evaluation is a genuine field deployment with eight therapist–PwD dyads (not a lab proxy); the authors report a technical audit of generated plans and suggestion frequencies; and the limitations section is candid, including the attribution confound and multiple-comparison risk. The main caveat is that the evidence supports system-level utility, not the specific causal contribution of the two核心技

major comments (3)
  1. [§5.2, §6.1–6.2, §8.6] The two headline numbers compare conditions that differ on several dimensions at once. In planning, the baseline (§5.2, App. A) requires authoring all questions/notes de novo while RemiAssist supplies an editable AI draft — a time reduction when editing a draft versus authoring from scratch is close to expected regardless of the Memory Graph's structure. In facilitation, the baseline shows only static therapist-authored questions while RemiAssist streams ~167 guiding questions, ~36 photo suggestions, and ~8 sensitivity prompts per session (Table 2); no non-context-aware or equal-information prompt condition was tested. The authors concede this in §8.6, but the abstract, §1 contributions, and §6 still frame the effects around the two named techniques. The fix is textual and proportionate: reframe claims as pertaining to the system-as-a-whole versus current practice (which is a legitimate
  2. [§6.2 (Conversation Duration)] Duration is treated as a positive outcome (54% increase, t6=2.55, dz=0.96), but longer sessions are not inherently better in RT: they could reflect prompt-stream-driven prolongation, therapist pacing, or PwD fatigue tolerance, and perceived conversation quality did not differ (Z=-1.00, p=.32). Since PwD self-report was not collected for ethical reasons (§8.1), the paper leans on duration plus therapist observation as the engagement signal. The authors should justify duration as a desirable endpoint (e.g., citing RT practice norms on session length), report its distribution per dyad rather than only the mean, and temper the abstract/§1 phrasing that presents the increase as an unqualified benefit.
  3. [§7 (Technical Evaluation)] The plan-quality audit (91%/90%/93% accuracy) rests on one therapist annotating errors with a second 'reviewing the annotations for reliability' — no inter-rater agreement statistic is reported, and it is unclear whether the reviewer independently re-annotated or merely adjudicated. Likewise, the blinded pre/post-edit usefulness ratings (5.8→6.4) come from two therapists with no agreement measure. Given that these numbers support the 'efficiency without quality loss' claim in §6.1, the paper should report an IRR statistic (e.g., Cohen's kappa or ICC) or describe the adjudication procedure precisely, and state whether accuracy was computed per element, per plan, or pooled.
minor comments (7)
  1. [§5.4, Table 3, Figure 6] With N=8 (N=7 for facilitation) and many uncorrected comparisons (Table 3), several p<.05 results would not survive even a mild correction. §5.4 and §8.6 flag this, but Figure 6 and Table 3 still use asterisks that invite confirmatory reading. Consider adding a note to the figure/table captions that tests are exploratory, or report an FDR-adjusted view alongside.
  2. [§4.4.2, §4.4.4] The scoring weights (W_rel=W_div=1, W_neg=100), the two-consecutive-detection hysteresis, and the 10s/20s refresh cadences were tuned with only two formative-study therapists (F1, F5). A brief sensitivity discussion (e.g., how ranking changes for moderate W_neg, or what fraction of candidates the W_neg=100 penalty suppresses) would help readers judge robustness; at minimum, note that these are heuristic settings not validated at scale.
  3. [Table 3, §6.2] Table 3 relabels NASA-TLX subscales as 'Low Mental Demand', etc., without stating the original item wording or direction; readers must infer that higher = lower demand. Please state the exact items and scale anchors. Also, the agency test statistic differs between §6.2 (Z=-0.48) and Table 3 (Z=-0.45) — please reconcile.
  4. [§6 (dropout handling)] The facilitation-phase dropout (T3–P3) is mentioned, but the analysis should state explicitly that all facilitation statistics are N=7 and discuss whether P3's withdrawal could correlate with condition (e.g., which system order the dyad had).
  5. [Front matter, Figure 1] 'Univeristy of Stuttgart' (author block); 'Context-A ware' hyphenation artifacts in Figure 1 caption, §1, and §4; Figure 1 caption runs into body text ('Figure 1:RemiAssistsupportstherapist-in-the-loopphoto-based...'). Please proof the camera-ready formatting.
  6. [§2.1] Reference [47] (Rememo, arXiv 2026) is highly concurrent work on an AI-in-the-loop therapist tool for dementia reminiscence. A sentence or two differentiating RemiAssist's contributions (graph-structured planning, real-time sensitivity handling, field deployment with PwD dyads) from Rememo's would clarify the novelty claim.
  7. [Appendix B, Table 1] The theme-distinguishability metric (within- vs between-theme edge weight ratio of 1.61) is computed on the same graph used to define the themes via Leiden clustering, so some separation is expected by construction. A baseline comparison (e.g., ratio under random theme assignment) would make this number interpretable.

Circularity Check

0 steps flagged · score 0.0 of 10

No derivation-by-construction: field-study endpoints are measured outcomes against a manual baseline, not algebraic restatements of fitted inputs.

full rationale

RemiAssist is an HCI systems paper whose load-bearing claims are empirical associations from a within-subject field study (planning time 7.6 vs 13.5 min; conversation duration 21.6 vs 14.0 min; NASA-TLX and Likert ratings), not first-principles predictions. The Memory Graph is a construction pipeline (context extraction → photo association → Leiden clustering → theme labeling) that produces editable drafts; plan quality is then checked by independent therapist annotation and pre/post-edit ratings, not asserted by definition. The Context-Aware Guiding Strategy uses a multi-objective score with therapist-tuned weights (W_rel=W_div=1, W_neg=100) and a two-detection sensitivity threshold—these are engineering hyperparameters that affect suggestion content, not parameters fitted to the primary endpoints and then re-reported as predictions. Self-citations in Related Work point to prior systems by overlapping authors as design context; none supply a uniqueness theorem or ansatz that forces the study results. Condition asymmetry and failure to isolate context-awareness (flagged by the authors in §8.6) are experimental confounds, not circular reductions. No step reduces a claimed result to its inputs by construction.

Assumptions & free parameters 3 free parameters · 5 assumptions · 2 invented entities

Load-bearing content is a designed interactive system plus a small clinical field comparison, not a formal derivation. Claims rest on standard HCI/clinical assumptions (PRT benefits, therapist centrality), engineering choices (VLM extractions, graph clustering, hand-tuned ranking weights and alert hysteresis), and the premise that the manual baseline isolates the value of the proposed techniques. Invented constructs are system abstractions (Memory Graph; suggestion taxonomy), not new physical entities.

free parameters (3)
  • Suggestion ranking weights W_rel, W_div, W_neg = W_rel=W_div=1, W_neg=100
    Multi-objective score S = W_rel S_rel + W_div S_div − W_neg S_neg; weights set by iterative testing with two formative therapists until ‘high overall quality,’ not cross-validated on held-out sessions.
  • Sensitivity alert hysteresis threshold = 2 consecutive detections (eval every 20s)
    Distress/off-topic flags require consecutive positive detections; one/two/three tested with two therapists; two chosen as best balance.
  • Suggestion refresh cadence and list sizes = 6 candidates/10s → top 3 questions; top 3 photos/20s
    Generation/curation rates and top-k chosen via iterative therapist testing, directly shaping real-time UX load.
assumptions (5)
  • domain assumption Photo-based reminiscence therapy is an effective psychosocial intervention for PwD when therapists prepare theme-centered materials and facilitate adaptively and ethically.
    Stated from clinical literature in §1 and formative study; underpins why planning/facilitation support matters.
  • domain assumption Autobiographical memory is usefully modeled as hierarchical themes → events/photos → shared episodic contexts (time, place, people, object, activity).
    §4.3 cites Conway et al.; justifies Memory Graph topology versus pure timelines.
  • domain assumption Vision-language and embedding models can extract and score photo/conversation context well enough for therapist-facing drafts and glanceable prompts (with human edit).
    Operational premise of construction and guiding pipelines (§4.3–4.4); accuracy audited post hoc ~90%+ on plans.
  • ad hoc to paper A manual gallery+notes baseline reflecting current therapist practice is an adequate control for claiming system benefits in planning and facilitation.
    §5.2–5.3 comparison design; authors later note it does not isolate context-awareness from information quantity (§8.6).
  • domain assumption Family-curated photo sets and pre-listed distressing topics sufficiently bound ethical risk for early-stage PwD sessions.
    §4.6, §5.2, §5.5 safety protocol; central to claiming safe sensitive-situation support.
invented entities (2)
  • Memory Graph (Themes → Photos → Context nodes with overlap edges)
    purpose: Structure personal photo archives for theme-centered PRT plan generation and retrieval.
    Paper-specific hierarchical graph instantiation combining face clustering, semantic merge, Leiden communities, and VLM labeling—not a previously standardized clinical artifact.
  • Context-Aware Guiding Strategy (guiding questions, relevant photos, sensitivity prompts)
    purpose: Real-time therapist suggestions conditioned on dialogue plus memory profile and distress list.
    Named system mechanism with paper-specific scoring, thresholds, and UI behavior evaluated in the field study.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RemiAssist: A Therapist-Supporting System for Photo-Based Reminiscence Therapy in Dementia Care." pith.science (2026). https://pith.science/paper/B4QWHT62

@misc{pith2026260724536,
  author       = {Pith},
  title        = {Pith review of: RemiAssist: A Therapist-Supporting System for Photo-Based Reminiscence Therapy in Dementia Care},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B4QWHT62}},
  note         = {Machine review of arXiv:2607.24536}
}
read the original abstract

Despite growing interest in applying AI to photo-based reminiscence therapy (PRT) for people with dementia (PwD), existing systems primarily focus on PwD-AI interaction and often overlook therapists' critical role in practical PRT delivery. We present RemiAssist, a system that supports therapist-in-the-loop PRT through AI-assisted planning and real-time facilitation. RemiAssist incorporates two core techniques: (1) a Memory Graph, which organizes key life events from a PwD's photo collection into a hierarchical graph to support theme-centered intervention planning; and (2) a Context-Aware Guiding Strategy, which provides real-time suggestions to help therapists guide reminiscence conversations and respond to sensitive situations. A field study with eight therapist-PwD dyads suggests that RemiAssist was associated with a 44% improvement in planning efficiency, a 54% increase in conversation duration, and timely support for handling sensitive situations. We highlight opportunities for AI systems to empower therapists and enable more personalized reminiscence therapy in dementia care.

Figures

Figures reproduced from arXiv: 2607.24536 by the authors.

Figure 1
Figure 1. RemiAssist supports therapist-in-the-loop photo-based reminiscence therapy in dementia care. It incorporates two core techniques: (A) a Memory Graph, which organizes key life events and their contextual relationships in a hierarchical graph; and (B) a Context-Aware Guiding Strategy, which provides real-time suggestions to help therapists facilitate reminiscence conversations and respond to sensitive situations. Abst… view at source ↗
Figure 2
Figure 2. RemiAssist supports therapists in planning, facilitating, and reflecting on each PRT session. (A) During planning, RemiAssist generates theme-centered session plans, which therapists can further revise based on their professional judgment. (B) During facilitation, RemiAssist provides therapists with three types of context-aware suggestions: (b1) guiding questions to sustain the conversation, (b2) relevant photos to … view at source ↗
Figure 4
Figure 4. illustrates the pipeline for constructing the Memory Graph. Given a photo collection curated for reminiscence use, in which each photo typically corresponds to a salient life event, the pipeline builds the Memory Graph in three stages: (1) context extraction, (2) photo association, and (3) theme generation. Context Extraction. In the first stage, RemiAssist extracts five types of contextual information from each pho… view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: RemiAssist generates three types of context-aware suggestions based on the real-time context and each PwD’s memory profile. 4.3.3 Intervention Plan Generation. After constructing the Memory Graph, RemiAssist maps each theme to a session plan automatically [PITH_FULL_I…
Figure 6
Figure 6. Figure 6: Therapists’ ratings for both systems (1 = strongly disagree, 7 = strongly agree). Error bars indicate standard errors. Asterisks indicate statistical significances based on the Wilcoxon signed-rank test (* denotes 𝒑 < .05) [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: RemiAssist reduced therapists’ session planning time and helped them sustain reminiscence conversations compared with the baseline. Statistical significance was assessed using paired t-tests. Error bars indicate standard errors. its ability to help them manage uncertai…
Figure 8
Figure 8. Figure 8: The baseline system is designed to reflect therapists’ manual PRT practice. During the preparation phase, therapists [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

80 extracted references · 5 linked inside Pith

  1. [1]

    2023.Visualization of time-oriented data

    Wolfgang Aigner, Silvia Miksch, Heidrun Schumann, and Christian Tominski. 2023.Visualization of time-oriented data. Springer Nature

    2023

  2. [2]

    Salvatore Andolina, Valeria Orso, Hendrik Schneider, Khalil Klouche, Tuukka Ruotsalo, Luciano Gamberini, and Giulio Jacucci. 2018. Investigating Proactive Search Support in Conversations. InProceedings of the 2018 Designing Interactive Systems Conference(Hong Kong, China)(DIS ’18). Association for Computing Machinery, New York, NY, USA, 1295–1307. doi:10....

    arXiv 2018

  3. [3]

    Andrea Baumann, Peter Shaw, Ludwig Trotter, Sarah Clinch, and Nigel Davies

  4. [4]

    Mariona Carós, Maite Garolera, Petia Radeva, and Xavier Giro-i Nieto. 2020. Automatic Reminiscence Therapy for Dementia. InProceedings of the 2020 Inter- national Conference on Multimedia Retrieval(Dublin, Ireland)(ICMR ’20). Asso- ciation for Computing Machinery, New York, NY, USA, 383–387. doi:10.1145/ 3372278.3391927

    arXiv 2020

  5. [5]

    Xiangxiang Chu, Limeng Qiao, Xinyu Zhang, Shuang Xu, Fei Wei, Yang Yang, Xiaofei Sun, Yiming Hu, Xinyang Lin, Bo Zhang, et al. 2024. Mobilevlm v2: Faster and stronger baseline for vision language model.arXiv preprint arXiv:2402.03766 (2024)

    arXiv 2024

  6. [6]

    Jennifer Chu-Carroll and Michael K Brown. 1997. Tracking initiative in col- laborative dialogue interactions. In35th Annual Meeting of the Association for Computational Linguistics and 8th Conference of the European Chapter of the Association for Computational Linguistics. 262–270

    1997

  7. [7]

    Victoria Clarke and Virginia Braun. 2017. Thematic analysis.The journal of positive psychology12, 3 (2017), 297–298

    2017

  8. [8]

    Martin A Conway. 2005. Memory and the self.Journal of memory and language 53, 4 (2005), 594–628

    2005

Show all 80 references
  1. [9]

    Martin A Conway and Debra A Bekerian. 1987. Organization in autobiographical memory.Memory & cognition15, 2 (1987), 119–132

    1987

  2. [10]

    Martin A Conway and Christopher W Pleydell-Pearce. 2000. The construction of autobiographical memories in the self-memory system.Psychological review107, 2 (2000), 261

    2000

  3. [11]

    Dagoberto Cruz-Sandoval, Arturo Morales-Tellez, Eduardo Benitez Sandoval, and Jesus Favela. 2020. A social robot as therapy facilitator in interventions to deal with dementia-related behavioral symptoms. InProceedings of the 2020 ACM/IEEE international conference on human-robo...

    2020

  4. [12]

    Pearl Ed G Cuevas, Patricia M Davidson, Joylyn L Mejilla, and Tamar W Rodney

  5. [13]

    Valdemar Danry, Eli Villa, Samantha Chan, and Pattie Maes. 2025. ReLive: Walk- ing into Virtual Reality Spaces from Video Recordings of One’s Past Can In- crease the Experiential Detail and Affect of Autobiographical Memories.IEEE Transactions on Visualization and Computer Gra...

    2025

  6. [14]

    Iván De-Rosende-Celeiro, Virginia Francisco-Gilmartín, Susana Bautista-Blasco, and Adriana Ávila-Álvarez. 2024. Co-creation and evaluation of an app to support reminiscence therapy interventions for older people with dementia.Digital Health 10 (2024), 20552076241261849

    2024

  7. [15]

    Martin Ester, Hans-Peter Kriegel, Jörg Sander, Xiaowei Xu, et al. 1996. A density- based algorithm for discovering clusters in large spatial databases with noise. In kdd, Vol. 96. 226–231

    1996

  8. [16]

    Guanghui Fu, Qing Zhao, Jianqiang Li, Dan Luo, Changwei Song, Wei Zhai, Shuo Liu, Fan Wang, Yan Wang, Lijuan Cheng, et al. 2023. Enhancing psychological counseling with large language model: A multifaceted decision-support system for non-professionals.arXiv preprint arXiv:2308...

    2023 arXiv

  9. [17]

    Siwei Fu, Hao Dong, Weiwei Cui, Jian Zhao, and Huamin Qu. 2017. How do ances- tral traits shape family trees over generations?IEEE transactions on visualization and computer graphics24, 1 (2017), 205–214

    2017

  10. [18]

    Sandra G Hart and Lowell E Staveland. 1988. Development of NASA-TLX (Task Load Index): Results of empirical and theoretical research. InAdvances in psy- chology. Vol. 52. Elsevier, 139–183

    1988

  11. [19]

    Steve Hodges, Lyndsay Williams, Emma Berry, Shahram Izadi, James Srinivasan, Alex Butler, Gavin Smyth, Narinder Kapur, and Ken Wood. 2006. SenseCam: A retrospective memory aid. InInternational conference on ubiquitous computing. Springer, 177–193

    2006

  12. [20]

    Junyuan Hong, Wenqing Zheng, Han Meng, Siqi Liang, Anqing Chen, Hiroko H Dodge, Jiayu Zhou, and Zhangyang Wang. 2024. A-conect: Designing ai-based conversational chatbot for early dementia intervention. InICLR 2024 Workshop on Large Language Model (LLM) Agents

    2024

  13. [21]

    Long-Jing Hsu, Alex Foster, Selma Sabanovic, and Chia-Fang Chung. 2025. Design- ing with Dynamics: Reflections on Co-design Workshops Between People Living with Dementia and Their Care Partners. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CH...

    2025

  14. [23]

    Hui-Chuan Huang, Yu-Ting Chen, Pin-Yuan Chen, Sophia Huey-Lan Hu, Fang Liu, Ying-Ling Kuo, and Hsiao-Yean Chiu. 2015. Reminiscence therapy improves cognitive functions and reduces depressive symptoms in elderly people with dementia: a meta-analysis of randomized controlled tri...

    2015

  15. [24]

    Weidong Huang, Peter Eades, and Seok-Hee Hong. 2009. Measuring effectiveness of graph visualizations: A cognitive load perspective.Information Visualization 8, 3 (2009), 139–152

    2009

  16. [25]

    Mina Huh, Zihui Xue, Ujjaini Das, Kumar Ashutosh, Kristen Grauman, and Amy Pavel. 2025. Vid2Coach: Transforming How-To Videos into Task Assistants. In Proceedings of the 38th Annual ACM Symposium on User Interface Software and Technology (UIST ’25). Association for Computing M...

    2025

  17. [26]

    Lei Jiang, Panote Siriaraya, Dongeun Choi, and Noriaki Kuwahara. 2021. A library of old photos supporting conversation of two generations serving reminiscence therapy.Frontiers in Psychology12 (2021), 704236

    2021

  18. [27]

    Yucheng Jin, Wanling Cai, Li Chen, Yizhe Zhang, Gavin Doherty, and Tonglin Jiang. 2024. Exploring the Design of Generative AI in Supporting Music-based Reminiscence for Older Adults. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems(Honolulu, HI, U...

    2024

  19. [28]

    Jacqueline Kindell, John Keady, Karen Sage, and Ray Wilkinson. 2017. Everyday conversation in dementia: A review of the literature to inform research and practice.International journal of language & communication disorders52, 4 (2017), 392–406

    2017

  20. [29]

    Noriaki Kuwahara, Shinji Abe, Kiyoshi Yasuda, and Kazuhiro Kuwabara. 2006. Networked reminiscence therapy for individuals with dementia by using photo and video sharing. InProceedings of the 8th international ACM SIGACCESS con- ference on Computers and accessibility. 125–132

    2006

  21. [30]

    Amanda Lazar, Hilaire Thompson, and George Demiris. 2014. A systematic review of the use of technology for reminiscence therapy.Health education & behavior41, 1_suppl (2014), 51S–61S

    2014

  22. [31]

    Matthew L Lee and Anind K Dey. 2007. Providing good memory cues for people with episodic memory impairment. InProceedings of the 9th international ACM SIGACCESS conference on Computers and accessibility. 131–138

    2007

  23. [32]

    Matan Levy, Rami Ben-Ari, Nir Darshan, and Dani Lischinski. 2023. Chatting makes perfect: Chat-based image retrieval.Advances in Neural Information Processing Systems36 (2023), 61437–61449

    2023

  24. [33]

    Jiahao Nick Li, Zhuohao Zhang, and Jiaju Ma. 2025. Omniquery: Contextually augmenting captured multimodal memories to enable personal question answer- ing. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. 1–20

    2025

  25. [34]

    Ruohao Li, Jiawei Li, Jia Sun, Zhiqing Wu, Zisu Li, Ziyan Wang, Ge Lin, and Mingming Fan. 2025. RemVerse: Supporting Reminiscence Activities for Older Adults through AI-Assisted Virtual Reality.Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.9, 3, Article 103 (Sept. 2025)...

    2025 doi

  26. [35]

    Bruce" Liu, Vladimir Kirilyuk, Xiuxiu Yuan, Alex Olwal, Peggy Chi, Xiang

    Xingyu "Bruce" Liu, Vladimir Kirilyuk, Xiuxiu Yuan, Alex Olwal, Peggy Chi, Xiang "Anthony" Chen, and Ruofei Du. 2023. Visual Captions: Augmenting Verbal Communication with On-the-fly Visuals. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems(Hambur...

    2023

  27. [36]

    Chan, Wazeer Zulfikar, Scott Ren, Christine Xu, and Pattie Maes

    Natasha Maniar, Samantha W.T. Chan, Wazeer Zulfikar, Scott Ren, Christine Xu, and Pattie Maes. 2025. MemPal: Leveraging Multimodal AI and LLMs for Voice-Activated Object Retrieval in Homes of Older Adults. InProceedings of the 30th International Conference on Intelligent User ...

    2025 doi

  28. [37]

    2024.Counseling and therapy skills

    David G Martin and Edward A Johnson. 2024.Counseling and therapy skills. Waveland Press

    2024

  29. [38]

    Liane-Marina Meßmer, Patrizia Held, Alexander Bejan, Christophe Kunze, and Christoph Reich. 2025. Designing a dementia caregiver-centered recommendation and context-aware AI media analytical application: a comprehensive system requirements analysis.International Journal of Hum...

    2025

  30. [39]

    Victor Morales-de Jesús, Helena Gómez-Adorno, María Somodevilla-García, and Darnes Vilariño. 2021. Conversational system as assistant tool in reminiscence therapy for people with early-stage of alzheimer’s. InHealthcare, Vol. 9. MDPI, RemiAssist UIST ’26, November 02–05, 2026,...

    2021

  31. [40]

    Hiromu Ogawa and Pattie Maes. 2020. Smartwatch-based topic suggestions to enrich casual conversations in awkward encounters. InProceedings of the 2020 ACM International Symposium on Wearable Computers(Virtual Event, Mexico) (ISWC ’20). Association for Computing Machinery, New ...

    2020

  32. [41]

    World Health Organization et al. 2021. Global status report on the public health response to dementia. (2021)

    2021

  33. [42]

    RAJA PARASURAMAN, PAMELA M GREENWOOD, JAMES V HAXBY, and CHERYL L GRADY. 1992. Visuospatial attention in dementia of the Alzheimer type.Brain115, 3 (1992), 711–733

    1992

  34. [43]

    Kyongok Park, Seonhye Lee, JeongEun Yang, Taekwon Song, and Gwi-Ryung Son Hong. 2019. A systematic review and meta-analysis on the effect of reminiscence therapy for people with dementia.International psychogeriatrics31, 11 (2019), 1581–1597

    2019

  35. [44]

    Pat Pataranutaporn, Chayapatr Archiwaranguprok, Samantha WT Chan, Eliza- beth Loftus, and Pattie Maes. 2025. Synthetic human memories: AI-edited images and videos can implant false memories and distort recollection. InProceedings of the 2025 CHI Conference on Human Factors in ...

    2025

  36. [45]

    S Tejaswi Peesapati, Victoria Schwanda, Johnathon Schultz, Matt Lepage, So-yae Jeong, and Dan Cosley. 2010. Pensieve: supporting everyday reminiscence. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. 2027–2036

    2010

  37. [46]

    Mohammad Rashidujjaman Rifat, Reem Ayad, Ashratuz Zavin Asha, Bingjian Huang, Selin Okman, Dina Sabie, Hasan Shahid Ferdous, Robert Soden, and Syed Ishtiaque Ahmed. 2024. Cohabitant: The design, implementation, and evaluation of a virtual reality application for interfaith lea...

    2024

  38. [47]

    Celeste Seah, Yoke Chuan Lee, Jung-Joo Lee, Ching-Chiuan Yen, and Clement Zheng. 2026. Rememo: A Research-through-Design Inquiry Towards an AI-in-the- loop Therapist’s Tool for Dementia Reminiscence.arXiv preprint arXiv:2602.17083 (2026)

    2026

  39. [48]

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

  40. [49]

    The Benevolent Society. 2005. Reminiscing Handbook

    2005

  41. [50]

    Ponnusamy Subramaniam and Bob Woods. 2012. The impact of individual reminiscence therapy for people with dementia: systematic review.Expert review of neurotherapeutics12, 5 (2012), 545–555

    2012

  42. [51]

    Jingwei Sun, Zhongyue Zhang, Mengyang Wang, Nianlong Li, Zhangwei Lu, Yan Xiang, Liuxin Zhang, Yu Zhang, Qianying Wang, and Mingming Fan. 2025. Chorus of the past: Toward designing a multi-agent conversational reminiscence system with digital artifacts for older adults. InProc...

    2025

  43. [52]

    Yuling Sun, Xiaojuan Ma, Silvia Lindtner, and Liang He. 2023. Care Workers’ Wellbeing in Data-Driven Healthcare Workplace: Identity, Agency, and Social Justice.Proceedings of the ACM on Human-Computer Interaction7, CSCW2 (2023), 1–29

    2023

  44. [53]

    Yuling Sun, Zhennan Yi, Xiaojuan Ma, Junyan Mao, and Xin Tong. 2024. Technology-Mediated Non-pharmacological Interventions for Dementia: Needs for and Challenges in Professional, Personalized and Multi-Stakeholder Collabora- tive Interventions. InProceedings of the 2024 CHI Co...

    2024

  45. [54]

    Vincent A Traag, Ludo Waltman, and Nees Jan Van Eck. 2019. From Louvain to Leiden: guaranteeing well-connected communities.Scientific reports9, 1 (2019), 5233

    2019

  46. [55]

    playing around

    Hsin-yi Sandy Tsai, Ruth Shillair, and Shelia R Cotten. 2017. Social support and “playing around” an examination of how older adults acquire digital literacy with tablet computers.Journal of Applied Gerontology36, 1 (2017), 29–55

    2017

  47. [56]

    Endel Tulving. 1984. Precis of elements of episodic memory.Behavioral and Brain Sciences7, 2 (1984), 223–238

    1984

  48. [57]

    Jindu Wang, Runze Cai, Shuchang Xu, Tianrui Hu, Huamin Qu, Shengdong Zhao, and Lin-Ping Yuan. 2026. Wearable AR for Restorative Breaks: How Interactive Narrative Experiences Support Relaxation for Young People. InProceedings of the 2026 CHI Conference on Human Factors in Compu...

    2026

  49. [58]

    Liyuan Wang, Xingxing Zhang, Hang Su, and Jun Zhu. 2024. A comprehensive survey of continual learning: Theory, method and application.IEEE transactions on pattern analysis and machine intelligence46, 8 (2024), 5362–5383

    2024

  50. [59]

    Peng Wang, Shuai Bai, Sinan Tan, Shijie Wang, Zhihao Fan, Jinze Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, et al. 2024. Qwen2-vl: Enhancing vision-language model’s perception of the world at any resolution.arXiv preprint arXiv:2409.12191(2024)

    2024 arXiv

  51. [60]

    Xin Wang, Juan Li, Tianyi Liang, Wordh Ul Hasan, Kimia Tuz Zaman, Yang Du, Bo Xie, Cui Tao, et al. 2024. Promoting personalized reminiscence among cognitively intact older adults through an AI-driven interactive multimodal photo album: Development and usability study.JMIR agin...

    2024

  52. [61]

    Steve Whittaker, Ofer Bergman, and Paul Clough. 2010. Easy on that trigger dad: a study of long term family photo retrieval.Personal and Ubiquitous Computing 14, 1 (2010), 31–43

    2010

  53. [62]

    Bob Woods, Laura O’Philbin, Emma M Farrell, Aimee E Spector, and Martin Orrell. 2018. Reminiscence therapy for dementia.Cochrane database of systematic reviews3 (2018)

    2018

  54. [63]

    Robert F Woolson. 2007. Wilcoxon signed-rank test.Wiley encyclopedia of clinical trials(2007), 1–3

    2007

  55. [64]

    World Health Organization. 2021. Closing the leadership gap: gender equity and leadership in the global health and care workforce. https://www.who.int/ publications/i/item/9789240025905

    2021

  56. [65]

    World Health Organization. 2025. Dementia. https://www.who.int/news-room/ fact-sheets/detail/dementia. accessed 2026-03-23

    2025

  57. [66]

    Dongmei Wu, Taolin Chen, Xiaoqi Huang, Lizhou Chen, Yuchuan Yue, Hao Yang, Xiuying Hu, and Qiyong Gong. 2020. The role of old photos in reminiscence therapy in elderly women with depressive symptoms: a functional magnetic resonance imaging study.Biological Research for Nursing...

    2020

  58. [67]

    Haijun Xia, Tony Wang, Aditya Gunturu, Peiling Jiang, William Duan, and Xiaoshuo Yao. 2023. CrossTalk: Intelligent Substrates for Language-Oriented Interaction in Video-Based Communication and Collaboration. InProceedings of the 36th Annual ACM Symposium on User Interface Soft...

    2023

  59. [68]

    Shuchang Xu, Chang Chen, Zichen Liu, Xiaofu Jin, Lin-Ping Yuan, Yukang Yan, and Huamin Qu. 2024. Memory Reviver: Supporting Photo-Collection Reminis- cence for People with Visual Impairment via a Proactive Chatbot. InProceedings of the 37th Annual ACM Symposium on User Interfa...

    2024

  60. [69]

    Shuchang Xu, Xiaofu Jin, Gaurav Jain, Wenshuo Zhang, Huamin Qu, Brian A Smith, and Yukang Yan. 2026. Sonic Stage: Automatically Generating an Interac- tive Spatial Soundscape to Facilitate Dialogue Video Comprehension for Blind and Low Vision Viewers. InProceedings of the Exte...

    2026

  61. [70]

    Shuchang Xu, Xiaofu Jin, Huamin Qu, and Yukang Yan. 2025. DanmuA11y: Making Time-Synced On-Screen Video Comments (Danmu) Accessible to Blind and Low Vision Users via Multi-Viewer Audio Discussions. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. 1–22

    2025

  62. [71]

    Shuchang Xu, Xiaofu Jin, Wenshuo Zhang, Huamin Qu, and Yukang Yan. 2025. Branch Explorer: Leveraging Branching Narratives to Support Interactive 360 Video Viewing for Blind and Low Vision Users. InProceedings of the 38th Annual ACM Symposium on User Interface Software and Tech...

    2025

  63. [72]

    Anna Xygkou, Chee Siang Ang, Panote Siriaraya, Jonasz Piotr Kopecki, Alexan- dra Covaci, Eiman Kanjo, and Wan-Jou She. 2024. MindTalker: Navigating the Complexities of AI-Enhanced Social Engagement for People with Early-Stage De- mentia. InProceedings of the 2024 CHI Conferenc...

    2024

  64. [73]

    An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. 2025. Qwen3 technical report.arXiv preprint arXiv:2505.09388(2025)

    2025 arXiv

  65. [74]

    Wenshuo Zhang, Leixian Shen, Shuchang Xu, Jindu Wang, Jian Zhao, Huamin Qu, and Lin-Ping Yuan. 2025. NeuroSync: Intent-Aware Code-Based Problem Solving via Direct LLM Understanding Modification. InProceedings of the 38th Annual ACM Symposium on User Interface Software and Tech...

    2025

  66. [75]

    In-Town" and

    Xuechen Zhang, Changyang He, Peng Zhang, Hansu Gu, Ning Gu, Qi Shen, Zhan Hu, and Tun Lu. 2025. RemiHaven: Integrating "In-Town" and "Out-of- Town" Peers to Provide Personalized Reminiscence Support for Older Drifters. In Proceedings of the 2025 CHI Conference on Human Factors...

    2025

  67. [76]

    Yanzhao Zhang, Mingxin Li, Dingkun Long, Xin Zhang, Huan Lin, Baosong Yang, Pengjun Xie, An Yang, Dayiheng Liu, Junyang Lin, et al. 2025. Qwen3 embedding: Advancing text embedding and reranking through foundation models.arXiv preprint arXiv:2506.05176(2025)

    2025 arXiv

  68. [77]

    Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, and Yanlin Wang. 2024. Memo- rybank: Enhancing large language models with long-term memory. InProceedings of the AAAI conference on artificial intelligence, Vol. 38. 19724–19731

    2024

  69. [78]

    Wazeer Deen Zulfikar, Samantha Chan, and Pattie Maes. 2024. Memoro: Using Large Language Models to Realize a Concise Interface for Real-Time Memory Augmentation. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems(Honolulu, HI, USA)(CHI ’24). Associa...

    2024 doi

  70. [2020]

    Reminiscence therapy for older adults with Alzheimer’s disease: a literature review.International journal of mental health nursing29, 3 (2020), 364–371

    2020

  71. [2023]

    Human–AI collaboration enables more empathic conversations in text- based peer-to-peer mental health support.Nature Machine Intelligence5, 1 (2023), 46–57

    2023

  72. [2024]

    InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems(Honolulu, HI, USA)(CHI ’24)

    Mnemosyne - Supporting Reminiscence for Individuals with Dementia in Residential Care Settings. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems(Honolulu, HI, USA)(CHI ’24). Association for Computing Machinery, New York, NY, USA, Article 217, 17 p...

    2024

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.