Pith. sign in

REVIEW 4 major objections 6 minor 36 references

HeadRoom: Lightweight, Edge-deployable Pipeline for Adaptive Notification Routing

T0 review · 4 major / 6 minor · reviewed 2026-07-10 · grok-4.5

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

Pith's one-line read Under high perceptual load, routing wearable notifications to the sensory channel with more residual capacity reduces response time.

desk verdict Solid lightweight systems paper with a real high-load RT effect and open edge code; the proxy and scope are the soft spots, not a collapse of the claim. read the letter →

arxiv 2607.08083 v1 pith:TC7YXOSO submitted 2026-07-09 cs.HC

Dinithi Dissanayake , Prasanth Sasikumar , Suranga Nanayakkara This is my paper
classification cs.HC
keywords multimodalinteractionwearablesdisruptionchannelavailabilitynotificationroutingegocentricsensingedgedeploymentpredictivecoding
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

Wearables such as smart glasses can deliver notifications by sight or by sound, yet designers still lack a practical way to pick the right channel at the right moment. HeadRoom is a lightweight, edge-deployable pipeline that watches egocentric video and audio and estimates residual capacity in each sensory channel by treating next-step prediction error as a real-time proxy for load. In a controlled study, when overall perceptual demand was high, sending brief probes to the channel the model judged more available produced faster responses than sending them to the less available channel; under lower demand the difference largely disappeared. The full pipeline runs on ordinary wearable hardware in about 11 ms per step with a sub-megabyte model footprint. If the approach generalizes, future wearable and immersive systems could interrupt people less disruptively by matching output modality to moment-to-moment channel headroom.

What carries the argument

Channel availability, defined as one minus the normalized next-step prediction error of a frozen-MobileNet embedding MLP (vision) and a 31-dimensional MFCC/flux MLP (audio). Higher prediction error is treated as higher occupancy; a simple comparison of the two availability scores decides the routing target.

What would settle it

In a high-demand egocentric scene, if participants respond equally fast or faster when probes are deliberately sent to the higher-prediction-error channel than to the lower-prediction-error channel, the central routing claim is false.

Watch

Extended reading notes

Core claim

HeadRoom demonstrates that visual and auditory channel availability—estimated continuously as inverted, online-normalized prediction error of separate lightweight next-step predictors on egocentric streams—can be used for adaptive notification routing. Under high perceptual load, routing probes to the more available channel measurably reduces response time relative to routing them to the less available channel.

Load-bearing premise

That how poorly a simple next-moment predictor anticipates the next visual embedding or audio feature vector is a faithful real-time stand-in for residual capacity in that sensory channel.

Editorial extensions

If this is right

  • When perceptual demand is high, systems can reduce response cost by preferring the currently freer sensory channel.
  • The same continuous availability signal can also help decide when to interrupt, not only which modality to use.
  • Sub-megabyte models and ~11 ms on-device latency make real-time channel-aware routing practical on contemporary XR headsets.
  • Routing benefits are largest under elevated demand; under low demand the advantage shrinks toward chance.

Reading between the lines

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

  • The same prediction-error idea could be extended to haptics, giving a third routing option when both vision and audition are occupied.
  • Availability traces might serve as a continuous, sensor-free secondary measure of channel load in dual-task psychology experiments.
  • Whether the millisecond-scale detection gains survive for richer notifications that require interpretation or action remains an open, testable question.
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

4 major / 6 minor

Summary. HeadRoom estimates moment-to-moment visual and auditory channel availability from egocentric streams by treating next-step prediction error (frozen MobileNetV3-Small 576-d embeddings + MLP for vision; 31-d MFCC/RMS/flux/ZCR + MLP for audio) as a proxy for residual perceptual capacity, after online EMA z-score normalization and inversion (§3.1–3.2, Alg. 1). A routing module then selects the higher-availability channel (with occupancy threshold τ=0.3 and tie δ=0.05). A controlled probe-detection study (N=25, analyzed N=22) on three held-out Aria scenarios under Model / Inverse / Random routing finds no reliable pooled effect, but under the highest-demand Video 3, Model yields faster RT than Inverse (≈−114 ms, p=.021; LME β≈−0.13 to −0.18 on log-RT), with trial-level correlations between availability (and availability delta) and RT. Edge feasibility is shown on Meta Quest 3S (mean ~11 ms/step, 0.625 MB).

Significance. If the high-load result generalizes, the work supplies a practical, open-source, edge-deployable primitive for modality-aware notification routing on wearables and XR devices—addressing a real gap between Multiple Resource Theory and deployable systems. Strengths include the lightweight self-supervised design, explicit runtime/memory numbers on commodity XR hardware, open models/ONNX/Unity prototype, and a psychophysical evaluation with mixed-effects models and availability–RT correlations rather than only subjective load. The contribution is scoped carefully to detection under high demand and does not overclaim ecological superiority over random routing.

major comments (4)
  1. §5.2.1–5.2.2 and Tables 3–5: The central claim is supported only in Video 3 (between-subjects cells n=7–8). Pooled Model vs Inverse is non-significant (Wilcoxon p=.166; paired t p=.435), and Model is not significantly faster than Random in Video 3 (p=.209; LME still favors both over Inverse). With counterbalancing producing small cells and only one high-demand clip showing the effect, the evidence for adaptive routing as a general design primitive is thin. Either power the high-demand contrast adequately (within-subjects or larger N), pre-register the Video-3 focus, or substantially qualify the abstract/conclusion claim that currently reads more broadly than the data.
  2. §5.1 / Appendix A.2 (probe selection): Probe onsets are chosen only at moments where the model’s higher availability ≥0.3. This couples the experimental stimulus set to the same signal under test and can inflate Model–Inverse ranking differences by excluding low-confidence or near-tie moments. Report sensitivity analyses without the filter (or with τ varied), and clarify how many candidate bins were discarded; otherwise the RT difference and availability–RT correlations partly reflect selection rather than pure routing validity.
  3. §3.1–3.2 and §5.2.3: The load-bearing assumption is that next-frame/window prediction error (after normalization) indexes residual channel capacity. Supporting correlations with low-level features are weak (Appendix Table 7; strongest r=−0.309 motion vs visual availability), and trial-level r(avail, RT)≈−0.11 / r(Δavail, RT)≈−0.20 are modest. The paper needs either (a) a stronger external validation of the proxy (e.g., against dual-task cost or known load manipulations independent of the routing labels) or (b) explicit framing that the result validates relative ranking under this operationalization, not that prediction error equals true spare capacity. Without that, the interpretation of Model vs Inverse as ‘channel-aware routing’ remains under-constrained.
  4. §5.2 and Limitations: Visual probes are systematically faster than auditory ones (574 vs 636 ms), which the authors attribute to visual priming from continuous screen viewing. Combined with fixed upper-right probe location and pure detection (not interpretation/action), this limits claims about real notification routing. The manuscript already notes ecological limits; the major issue is that the abstract and contribution statements still present the result as evidence for adaptive notification routing in wearables. Tighten those statements to match the detection-task, high-load, Model-vs-Inverse scope, or add a richer secondary measure.
minor comments (6)
  1. Abstract and §1: N=25 is stated; analyzed N=22 after exclusion for the 80% training threshold. Report analyzed N consistently in the abstract.
  2. Eq. (1) and Appendix A.1: τ=0.3 and δ=0.05 are called ‘practical operating values’; post-hoc note that benefits weaken at τ≤0.2 is useful—move a brief sensitivity statement into the main text so readers see parameter dependence without only the appendix.
  3. Figure 3 / Tables 3–5: Provide error bars or CIs on the per-condition means and state whether means are participant-level or trial-level aggregates.
  4. §4.1: Live Aria streaming observations are qualitative and pointed to the website; a short quantitative summary (e.g., direction of availability shifts under abrupt sound vs high motion) in the main text would strengthen the feasibility narrative.
  5. Typos/clarity: ‘wihtout’ (§2), ‘Promemassist’ vs ‘Promemas-sist’ inconsistency with citation [22], and ‘either when the difference is small’ in the routing description could be tightened.
  6. NASA-TLX null result is informative; report effect sizes or Bayes factors if space allows so readers can judge evidence of absence vs underpowering.

Circularity Check

0 steps flagged · score 1.0 of 10

No load-bearing circularity: prediction-error availability is an independent sensory statistic validated against separate behavioral RT; minor self-citations are non-essential related work.

full rationale

The derivation chain is: (1) define channel availability as inverted, online-normalized next-step prediction error of lightweight self-supervised predictors on egocentric streams (Eq. after §3.1, Alg. 1); (2) train predictors only on Aria Everyday Activities (held-out scenarios reserved for study); (3) route probes by comparing the two availability scores and measure participant RT/detection on a separate psychophysical task. RT is never an input to the predictors, normalizer, or routing rule; thresholds τ=0.3 and δ=0.05 are stated a priori operating values, not fitted to the RT outcome. Trial-level correlations (r≈−0.11 / −0.20) and the Video-3 Model-vs-Inverse ΔRT are therefore external validations, not tautologies. Self-citations ([11], [24], [25]) appear only in related-work discussion of cognitive-load sensing and do not supply a uniqueness theorem, ansatz, or fitted parameter that forces the present result. Probe-time filtering by α≥0.3 is a design choice that may limit generalizability but does not make the ranking comparison circular by construction. The paper is therefore self-contained against its own behavioral benchmark; score remains at the minor-self-citation floor.

Assumptions & free parameters 5 free parameters · 4 assumptions · 1 invented entities

The central claim rests on treating next-step prediction error as residual channel capacity, on modality-specific resource pools, and on a handful of hand-set routing thresholds. No new physical entities are postulated; the invented construct is the operationalized availability score itself.

free parameters (5)
  • occupancy threshold τ = 0.3
    Hand-set to 0.3; probes only scheduled when availability ≥ τ; appendix notes benefits weaken if lowered to 0.2.
  • tie threshold δ = 0.05
    Hand-set to 0.05 for “either” routing when channels are close.
  • normalization warm-up length = 30 steps / 3 s
    3 s / 30 steps of neutral 0.5 availability before routing; EMA rates α_μ, α_σ and clip range c are implementation choices.
  • audio log-compression constant = 1000
    log(1+1000 e_t) applied only to audio errors before normalization.
  • availability floor ε = 0.05
    Hard floor at 0.05 to avoid zero availability.
assumptions (4)
  • domain assumption Higher next-step prediction error in a sensory stream implies greater processing demand and therefore lower residual capacity for new information (predictive-coding proxy).
    Stated in §3.1 and used to define α_m_t = 1 − normalized error; not independently validated outside the RT correlations.
  • domain assumption Visual and auditory attention draw on partly separate resource pools, so channel-specific availability is meaningful for routing (Multiple Resource Theory).
    Invoked in Related Work and system design; justifies separate predictors rather than a single load score.
  • ad hoc to paper A frozen MobileNetV3-Small embedding plus a two-layer MLP next-embedding predictor is a sufficient visual load sensor for real-time routing.
    Architecture and training choices in §3.2.1; alternatives not ablated against RT.
  • ad hoc to paper 31-d MFCC/RMS/flux/ZCR features plus an MLP next-window predictor is a sufficient auditory load sensor.
    §3.2.2; feature set chosen for compactness, not proven optimal for availability.
invented entities (1)
  • channel availability α_m_t = 1 − normalized prediction error independent evidence
    purpose: Scalar residual capacity used to choose visual vs auditory notification routing in real time.
    Defined by the authors as the operational construct; independent behavioral handle is the RT correlation and Model-vs-Inverse contrast, which is partial.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HeadRoom: Lightweight, Edge-deployable Pipeline for Adaptive Notification Routing." pith.science (2026). https://pith.science/paper/TC7YXOSO

@misc{pith2026260708083,
  author       = {Pith},
  title        = {Pith review of: HeadRoom: Lightweight, Edge-deployable Pipeline for Adaptive Notification Routing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TC7YXOSO}},
  note         = {Machine review of arXiv:2607.08083}
}
read the original abstract

Emerging wearables, such as smart glasses, can deliver notifications through multiple sensory channels, but there is still a limited understanding of how to choose the right channel at the right moment. We propose HeadRoom, a lightweight, edge-deployable pipeline that estimates the availability of visual and auditory channels in real time from egocentric video and audio. Our controlled user study (N=25) shows that, under high perceptual load, routing notifications to the more available channel reduces response time relative to routing them to the less available channel. This work opens up a new possibility for adaptive routing of notifications in wearable and immersive systems.

Figures

Figures reproduced from arXiv: 2607.08083 by the authors.

Figure 1
Figure 1. HeadRoom is a lightweight, low-latency pipeline that analyzes egocentric visual and auditory streams in real time to [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. HeadRoom system pipeline. Egocentric video and audio are processed by lightweight predictors, whose errors estimate [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Mean response times for Model, Random, and Inverse. The only significant difference appeared in video 3, where Model is faster than Inverse, suggesting that routing to the more available channel reduces response time under high demand. 5.2.2 HeadRoom Reduced Response Time Under High De￾mand. Video 3 showed the clearest significant condition-dependent pattern. A Kruskal–Wallis test revealed a significant difference a… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Visual probe - square in the upper-right corner, [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Participants completing the study in the headset [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [1]

    Riku Arakawa, Hiromu Yakura, and Mayank Goel. 2024. PrISM-Observer: In- tervention agent to help users perform everyday procedures sensed using a smartwatch. InProceedings of the 37th Annual ACM Symposium on User Interface Software and Technology. 1–16

    work page 2024

  2. [2]

    Alan D Baddeley. 2013. The concept of working memory. InModels of short-term memory. Psychology Press, 1–27

    work page 2013

  3. [3]

    Brian P Bailey and Shamsi T Iqbal. 2008. Understanding changes in mental work- load during execution of goal-directed tasks and its application for interruption management.ACM Transactions on Computer-Human Interaction (TOCHI)14, 4 (2008), 1–28

    work page 2008

  4. [4]

    Runze Cai, Nuwan Janaka, Hyeongcheol Kim, Yang Chen, Shengdong Zhao, Yun Huang, and David Hsu. 2025. AiGet: Transforming Everyday Moments into Hidden Knowledge Discovery with AI Assistance on Smart Glasses. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. 1–26

    work page 2025

  5. [5]

    Samantha WT Chan, Shardul Sapkota, Rebecca Mathews, Haimo Zhang, and Suranga Nanayakkara. 2020. Prompto: Investigating receptivity to prompts based on cognitive load from memory training conversational agent.Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies4, 4 (2020), 1–23

    work page 2020

  6. [6]

    Hyunsung Cho, Drew Edgar, David Lindlbauer, and Joseph O’Hagan. 2025. Eval- uating dynamic delivery of audio+ visual message notifications in xr. In2025 IEEE conference virtual reality and 3D user interfaces (VR). IEEE, 277–287

    work page 2025

  7. [7]

    Andy Clark. 2013. Whatever next? Predictive brains, situated agents, and the future of cognitive science.Behavioral and brain sciences36, 3 (2013), 181–204

    work page 2013

  8. [8]

    1990.Flow: The psychology of optimal experience

    Mihaly Csikszentmihalyi and Mihaly Csikzentmihaly. 1990.Flow: The psychology of optimal experience. Vol. 1990. Harper & Row New York

    work page 1990

Show all 36 references
  1. [9]

    Steven Davis and Paul Mermelstein. 1980. Comparison of parametric representa- tions for monosyllabic word recognition in continuously spoken sentences.IEEE transactions on acoustics, speech, and signal processing28, 4 (1980), 357–366

    1980

  2. [10]

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. Imagenet: A large-scale hierarchical image database. In2009 IEEE conference on computer vision and pattern recognition. Ieee, 248–255

    2009

  3. [11]

    Dinithi Dissanayake and Suranga Nanayakkara. 2025. Navigating the State of Cognitive Flow: Context-Aware AI Interventions for Effective Reasoning Support. arXiv preprint arXiv:2504.16021(2025)

    2025 arXiv

  4. [12]

    Karl Friston and Stefan Kiebel. 2009. Predictive coding under the free-energy principle.Philosophical transactions of the Royal Society B: Biological sciences364, 1521 (2009), 1211–1221

    2009

  5. [13]

    Violet Yinuo Han, Abena Boadi-Agyemang, Yuyu Lin, David Lindlbauer, and Alexandra Ion. 2023. Parametric Haptics: Versatile Geometry-based Tactile Feedback Devices. InProceedings of the 36th Annual ACM Symposium on User Interface Software and Technology(San Francisco, CA, USA)(...

    2023

  6. [14]

    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

  7. [15]

    Andrew Howard, Mark Sandler, Grace Chu, Liang-Chieh Chen, Bo Chen, Mingx- ing Tan, Weijun Wang, Yukun Zhu, Ruoming Pang, Vijay Vasudevan, et al. 2019. Searching for mobilenetv3. InProceedings of the IEEE/CVF international conference on computer vision. 1314–1324

    2019

  8. [16]

    1973.Attention and Effort

    Daniel Kahneman. 1973.Attention and Effort. Prentice-Hall, Englewood Cliffs, NJ

    1973

  9. [17]

    Zhaoyang Lv, Nicholas Charron, Pierre Moulon, Alexander Gamino, Cheng Peng, Chris Sweeney, Edward Miller, Huixuan Tang, Jeff Meissner, Jing Dong, et al

  10. [18]

    Aria everyday activities dataset.arXiv preprint arXiv:2402.13349(2024)

    2024 arXiv

  11. [19]

    James SP Macdonald and Nilli Lavie. 2011. Visual perceptual load induces inatten- tional deafness.Attention, Perception, & Psychophysics73, 6 (2011), 1780–1789

    2011

  12. [20]

    Helen J Neville and Donald Lawson. 1987. Attention to central and periph- eral visual space in a movement detection task: An event-related potential and behavioral study. I. Normal hearing adults.Brain research405, 2 (1987), 253–267

    1987

  13. [21]

    AM Parkes, N Coleman, et al . 1990. Route guidance systems: A comparison of methods of presenting directional information to the driver.Contemporary ergonomics22, 2 (1990), 480–485

    1990

  14. [22]

    Michael J Proulx, David J Brown, Achille Pasqualotto, and Peter Meijer. 2014. Multisensory perceptual learning and sensory substitution.Neuroscience & Biobehavioral Reviews41 (2014), 16–25

    2014

  15. [23]

    Kevin Pu, Ting Zhang, Naveen Sendhilnathan, Sebastian Freitag, Raj Sodhi, and Tanya R Jonker. 2025. Promemassist: Exploring timely proactive assistance through working memory modeling in multi-modal wearable devices. InProceed- ings of the 38th Annual ACM Symposium on User Int...

    2025

  16. [24]

    Rajesh PN Rao and Dana H Ballard. 1999. Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects.Nature neuroscience2, 1 (1999), 79–87

    1999

  17. [25]

    Prasanth Sasikumar, Ryo Hajika, Kunal Gupta, Tamil Selvan Gunasekaran, Yun Suen Pai, Huidong Bai, Suranga Nanayakkara, and Mark Billinghurst

  18. [26]

    In2024 IEEE Conference Virtual Reality and 3D User Interfaces (VR)

    A User Study on Sharing Physiological Cues in VR Assembly Tasks. In2024 IEEE Conference Virtual Reality and 3D User Interfaces (VR). 765–773. doi:10.1109/VR58804.2024.00096

    2024 doi

  19. [27]

    Prasanth Sasikumar, Yun Suen Pai, Huidong Bai, and Mark Billinghurst. 2022. PSCVR: Physiological Sensing in Collaborative Virtual Reality. In2022 IEEE Inter- national Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct). 663–666. doi:10.1109/ISMAR-Adjunct57072.2022.00137

    2022 doi

  20. [28]

    Issatay Tokmurziyev, Miguel Altamirano Cabrera, Muhammad Haris Khan, Yara Mahmoud, and Dzmitry Tsetserukou. 2026. LLM-Glasses: GenAI-driven Glasses with Haptic Feedback for Navigation of Visually Impaired People. arXiv:2503.16475 [cs.HC] https://arxiv.org/abs/2503.16475

    2026

  21. [29]

    Anne M Treisman and Alison Davies. 1973. Divided attention to ear and eye. Attention and performance IV(1973), 101–117

    1973

  22. [30]

    Robert B Welch and David H Warren. 1980. Immediate perceptual response to intersensory discrepancy.Psychological bulletin88, 3 (1980), 638

    1980

  23. [31]

    Shaoyue Wen, Songming Ping, Jialin Wang, Hai-Ning Liang, Xuhai Xu, and Yukang Yan. 2024. AdaptiveVoice: Cognitively adaptive voice interface for driv- ing assistance. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems. 1–18

    2024

  24. [32]

    Christopher D Wickens. 2002. Multiple resources and performance prediction. Theoretical issues in ergonomics science3, 2 (2002), 159–177

    2002

  25. [33]

    Christopher D Wickens. 2008. Multiple resources and mental workload.Human factors50, 3 (2008), 449–455

    2008

  26. [34]

    Christopher D Wickens, Diane L Sandry, and Michael Vidulich. 1983. Compati- bility and resource competition between modalities of input, central processing, and output.Human factors25, 2 (1983), 227–248

    1983

  27. [35]

    Sijia Zhao, Christopher A Brown, Lori L Holt, and Frederic Dick. 2022. Ro- bust and efficient online auditory psychophysics.Trends in hearing26 (2022), 23312165221118792

    2022

  28. [36]

    Jingyao Zheng, Xian Wang, Sven Mayer, and Lik-Hang Lee. 2026. Non-urgent Messages Do Not Jump into My Headset Suddenly! Adaptive Notification Design in Mixed Reality. arXiv:2603.05893 [cs.HC] https://arxiv.org/abs/2603.05893 HeadRoom: Lightweight, Edge-deployable Pipeline for ...

    2026

Pith tools

Reviewed July 10, 2026 · model on record in the stance chip above.