Pith. sign in

REVIEW 3 major objections 5 minor 39 references

Map as a By-product: Collective Landmark Mapping from IMU Data and User-provided Texts in Situated Tasks

T0 review · 3 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read Aggregating smartphone motion data and routine text notes from several workers can produce accurate indoor landmark maps—about 0.8 m average error in a small retail store and 1.9 m in an office—as a by-product of their normal tasks.

desk verdict Useful integration paper with a plausible feasibility demo; the cross-user alignment step is under-analyzed and has a real observability gap, but the work deserves a serious referee. read the letter →

arxiv 2509.03792 v1 pith:XGP3WYQJ submitted 2025-09-04 cs.HC

Ryo Yonetani , Kotaro Hara This is my paper · ORCID
classification cs.HC
keywords participatorysensingsemanticlandmarkmappinginertialnavigationindoorlocalizationlargelanguagemodelscoordinatealignmentcrowdsourcedsituatedtasks
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

The paper claims that indoor semantic maps can be generated as a by-product of work people already do: while a store employee notes "three cookies left" or an inspector types "Room 19 free," the phone's IMU records their path, an LLM extracts the landmark name, and a new alignment optimization fuses many such logs into a single map. The reported result is an average positional error of 0.8 m in a 70 m2 retail store and 1.9 m in a 300 m2 office, with nearly complete landmark coverage once 11 users' records are combined. If correct, this would give retailers and facility managers up-to-date, location-tagged semantic data without site surveys, cameras, or dedicated hardware.

What carries the argument

The load-bearing mechanism is the coordinate alignment objective. Per recording, the system optimizes a rigid transform (rotation R and translation t) applied to all landmarks in that recording's own coordinate frame; the objective is $L = \sum_{i,j} S_{ij} \|q_i - q_j\|^2$ with $q_i = R_{\Omega_i} p_i + t_{\Omega_i}$. $S_{ij}$ is a relatedness score between inferred landmark names, computed from text-embedding distances. The objective pulls semantically similar landmarks together across recordings while the rigid transforms keep each recording's internal geometry intact, so drift and unknown starting pose are corrected collectively rather than by an absolute reference. The system then takes median pos

What would settle it

Run the proposed coordinate alignment on simulated recording sets with known ground-truth transforms and a landmark-similarity graph that is disconnected or contains duplicate-name cross-links; if gradient descent from random initializations lands in visibly misaligned local minima with positional error far above the reported 0.8 m, while an exhaustive search over transforms finds a much better configuration, the central claim that aggregation alone fixes drift would be refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that a semantic landmark map of an indoor space can be produced as a by-product of everyday situated tasks, using only the motion signals already captured by a phone and the free text workers type while doing their jobs. The pipeline is: neural inertial navigation converts each user's IMU stream into a trajectory; a large language model reads the user's note and infers a landmark name; the position where the user stopped is attached to that name. Because each recording has its own origin and heading, the authors introduce a coordinate alignment optimization that estimates a rotation and translation per recording and minimizes $L = \sum S_{ij} \|q_i - q_j\|^2$, where S

Load-bearing premise

The load-bearing premise is that gradient descent on the alignment objective reliably reaches a global configuration that correctly couples the recordings; if it gets stuck in a local minimum or the landmark-similarity graph is disconnected or ambiguous, the fused map can be arbitrarily misaligned, and the paper offers only empirical evidence, not a proof, that this does not happen.

Editorial extensions

If this is right

  • A store or office can get a semantic map without a dedicated mapping pass: workers' routine restocking or inspection logs become the data source.
  • More participants systematically improve both positional accuracy and landmark coverage; at 11 records the system reaches near-perfect coverage, and reported error keeps dropping at 20 and 30 records.
  • A few rich recordings that visit many landmarks can anchor alignment, so even mixed participation levels are useful.
  • Dynamic semantic data (stock counts, room occupancy) are collected with timestamps in the same by-product pass, enabling location-aware chatbots.
  • Duplicate-name landmarks are the main known failure mode, and the paper shows that excluding non-unique names from the alignment step restores performance.

Reading between the lines

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

  • Editorial inference: the alignment objective is essentially a synchronization problem over the graph of landmark relatedness, so field robustness likely depends on that graph being connected; a practical system might monitor graph connectivity and report confidence, something the paper does not analyze.
  • Editorial inference: because the paper evaluates positional error by SVD-aligning the generated map to ground truth, the reported accuracy is a post-hoc measure; in a deployment without floor plans, users would experience relative consistency rather than absolute meters, so a useful next test is whether the maps support navigation without any ground-truth alignment.
  • Editorial inference: the same alignment machinery could extend to multi-floor settings by combining with barometric pressure (which the paper names as future work), but a floor-discriminating term would need to be added and tested.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes Collective Landmark Mapper, a system that builds indoor semantic landmark maps as a by-product of situated tasks. During routine activities such as restocking checklists in a retail store and room inspections in an office, users provide free-text notes on smartphones while IMU data are recorded. The system uses a neural inertial navigation model to reconstruct trajectories, an LLM to extract landmark names and semantic information from the text, and a coordinate-alignment optimization that rigidly transforms each user's recording so that landmarks with similar names are placed close together. The authors report a user study with 16 participants and 39/36 valid records, claiming average positional errors of 0.8 m in the retail store and 1.9 m in the office, nearly complete landmark coverage with 11 records, and improved accuracy as more records are aggregated. A simulation study varies the number of landmarks per recording, positional noise, and landmark duplication. The paper frames the contribution as an easy-to-deploy, privacy-preserving alternative to vision-based or infrastructure-dependent indoor mapping.

Significance. If the claims hold, the system would be a meaningful step toward low-cost, incidentally produced indoor semantic maps: it uses only IMU and free text, avoids cameras and site surveys, and leverages existing staff workflows. The work has real practical motivation, grounded in a formative interview, and the evaluation is carried out in actual retail and office environments rather than a lab. The system is also compared against two vision-based baselines, and the simulation explores several factors that a single user study cannot cover. The main intellectual contribution is the coordinate-alignment formulation, but its correctness and robustness are not yet established beyond the specific empirical settings tested. The paper does not exhibit circularity: the alignment objective is defined independently of the ground-truth maps, the LLM and embedding models are external, and evaluation uses measured ground truth. No code or data are shipped, which limits reproducibility but does not by itself invalidate the results.

major comments (3)
  1. [Sec. 4.3 (coordinate alignment optimization)] The optimization objective L = sum S_ij ||q_i - q_j||^2 is invariant to rotations of any recording that has fewer than two landmarks with non-negligible relatedness to landmarks in other recordings. Concretely, if exactly one landmark in a recording is matched to landmarks outside that recording, rotating the whole recording around that matched point leaves L essentially unchanged, so the positions of that recording's other landmarks are determined by initialization and gradient-descent trajectory, not by data. The paper gives no initialization scheme, no condition for identifiability of the rigid transforms, no analysis of local minima, and no report of how many of the 39/36 user-study recordings were underconstrained in this sense. This is load-bearing for the central claim that aggregating more users improves map accuracy. The authors should report the connectivity of the landmark-rel
  2. [Sec. 5.2/5.3, Fig. 8] The evaluation aligns the generated map to the ground-truth map using an optimal similarity transformation computed by SVD before measuring Euclidean error. This removes global translation, rotation, and uniform scaling, so the reported 0.8 m and 1.9 m errors are not absolute positioning errors; they can mask systematic global or rotational offsets. Because each recording's coordinate frame is arbitrary, some global alignment is necessary, but the authors should distinguish alignment error from residual error and, where possible, report results with a fixed registration as well. In addition, Fig. 8 shows only point estimates with no error bars or significance tests, despite being based on ten random record selections. The claimed monotonic improvement with increasing record count and the claimed superiority of user-provided texts over image captioning in the office task require variance
  3. [Sec. 6.1-6.2 (simulation setup)] The simulation samples landmark subsets uniformly at random from 30 unique IDs. In the 'few' condition with K=3 and up to 15 records, this uniform sampling makes it likely that most records share at least two landmarks with other records, so the simulation does not exercise the underconstrained-recording failure mode identified in Sec. 4.3. The paper does not report how many simulated records were underconstrained, nor does it vary the overlap structure to mimic realistic non-uniform visit patterns, such as users visiting disjoint areas of a store. The simulation therefore overstates robustness with respect to the main optimization risk. The authors should add a condition with deliberately sparse overlap, report the number of underconstrained records, and show the effect on positional error.
minor comments (5)
  1. [Sec. 5.4] The Discussion reports results for 20 and 30 aggregated records (0.63 m and 0.56 m in retail; 1.4 m and 1.2 m in office), but these points are not shown in Fig. 8, which appears to stop at 10-11 records. Please clarify whether these are from the same 39/36 records and add them to the figure or explain the discrepancy.
  2. [Sec. 4.3/4.4] Implementation details of the optimization are missing: learning rate, convergence criterion, number of iterations, initialization (e.g., identity transforms), and whether any restarts are used. Even a short paragraph would substantially improve reproducibility.
  3. [Sec. 5.3, Fig. 8] The x-axis label of Fig. 8 shows '# Records' with ticks 2,4,6,8,10, while the text says aggregation levels from 1 to 11 records. Please make the tick labels consistent with the protocol.
  4. [Sec. 5.1] Participants were recruited through personal contacts and word of mouth, and all were familiar with the environments. This is a convenience sample; the authors should state this limitation explicitly in Sec. 5.4 or the conclusion, and avoid generalizing beyond feasibility claims.
  5. [Sec. 6.3 and Fig. 11b] The simulation uses sigma ~ 0.5 based on user-study data, but Fig. 11b varies sigma and shows strong sensitivity. Please state the number of random seeds used for each configuration in the text or caption; the current text says five seeds but it is not clear whether this applies to every panel.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the landmark map is computed from IMU/text inputs and evaluated against external ground truth.

full rationale

The paper's central derivation chain is self-contained rather than circular. Landmark positions come from neural inertial navigation on IMU data and LLM-inferred names from user text; the coordinate alignment objective L = sum S_ij ||q_i - q_j||^2 is defined over these independently produced noisy inputs and per-recording rigid transforms, with no ground-truth map, fitted target, or final map appearing in the objective. The relatedness scores come from an external text-embedding model, and the evaluation uses manually annotated ground-truth shelf/room positions, with an SVD similarity transform used only to place the generated map into the ground-truth coordinate frame after optimization—not to fit the map's internal structure. The reported 0.8 m / 1.9 m errors and coverage values are therefore empirical outcomes, not quantities forced by construction. The simulation injects noise with sigma ~ 0.5 'based on the user study data,' but that calibration does not feed the central claim about the real user study, and the simulation is presented as a secondary systematic analysis. The only self-citation in the reference list ([32], conversational localization) is not load-bearing for the mapping claim, and no uniqueness theorem or ansatz is imported from the authors' prior work. The skeptic's concern about underconstrained recordings making the objective rotation-invariant is a genuine robustness/correctness risk, but it is not an instance of a prediction reducing to its own inputs; the paper's claims could be wrong for that reason without any circularity. Overall, no circular step was found.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

No new physical entities, forces, particles, or sensors are postulated. The 'map' and 'landmark' are aggregations of existing measurements and text. The main assumptions are about the accuracy of externally supplied neural inertial navigation and LLM components, the alignment optimization's behavior, and the evaluation protocol.

free parameters (3)
  • Noise scale sigma = 0.5 (standard deviation in meters)
    Set from observed variability in user-to-landmark distance in the user study; used only in simulation, not in the core user-study results.
  • LLM landmark category list for retail prompt = ten hand-written categories (Stationery, Beverages, Body care, Toy, Snacks, Books, etc.)
    The retail prompt constrains LLM classification to this closed set; accuracy of landmark naming depends on this choice.
  • Simulation sampling conditions K = few: 3-6; many: 12-15; mixed: 10% many
    Chosen to represent user-study workloads; drives coverage conclusions in Sec. 6.2.
assumptions (6)
  • domain assumption RoNIN neural inertial navigation reconstructs trajectories with sufficient accuracy for the alignment step.
    Sec. 4.1 uses RoNIN [37] without floorplan corrections; errors are treated as noise rather than modeled.
  • domain assumption A user's stationary position while typing text approximates the physical landmark position.
    Sec. 4.2 averages the estimated position over the detected stop; participants typed from variable distances (Sec. 5.4).
  • domain assumption LLM and text embeddings map free-form task notes to landmark names whose relatedness scores correspond to physical co-location.
    Sec. 4.2 and Sec. 4.4; misclassification is acknowledged in Sec. 5.4 for Alinamin V/Energy Drinks.
  • ad hoc to paper The alignment objective has a useful local optimum reachable by gradient descent.
    Sec. 4.3 states minimization "via gradient descent until convergence" with no proof or implementation details; correctness is inferred from user study and simulation outcomes.
  • domain assumption SVD least-squares alignment of predicted landmarks to ground truth with known correspondences is a fair evaluation protocol.
    Sec. 5.2 computes errors after optimal similarity transformation, which can hide consistent absolute offset or rotation errors.
  • domain assumption Gaussian noise with sigma ~0.5 models IMU drift and standing-position variability in simulation.
    Sec. 6.1 calibrates noise to user-study data; real drift is often non-Gaussian and grows over time.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Map as a By-product: Collective Landmark Mapping from IMU Data and User-provided Texts in Situated Tasks." pith.science (2026). https://pith.science/paper/XGP3WYQJ

@misc{pith2026250903792,
  author       = {Pith},
  title        = {Pith review of: Map as a By-product: Collective Landmark Mapping from IMU Data and User-provided Texts in Situated Tasks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XGP3WYQJ}},
  note         = {Machine review of arXiv:2509.03792}
}
read the original abstract

This paper presents Collective Landmark Mapper, a novel map-as-a-by-product system for generating semantic landmark maps of indoor environments. Consider users engaged in situated tasks that require them to navigate these environments and regularly take notes on their smartphones. Collective Landmark Mapper exploits the smartphone's IMU data and the user's free text input during these tasks to identify a set of landmarks encountered by the user. The identified landmarks are then aggregated across multiple users to generate a unified map representing the positions and semantic information of all landmarks. In developing the proposed system, we focused specifically on retail applications and conducted a formative interview with stakeholders to confirm their practical needs that motivate the map-as-a-byproduct approach. Our user study demonstrates the feasibility of the proposed system and its superior mapping performance in two different setups: creating a product availability map from restocking checklist tasks at a retail store and constructing a room usage map from office inspection tasks, further demonstrating the potential applicability to non-retail applications.

Figures

Figures reproduced from arXiv: 2509.03792 by the authors.

Figure 1
Figure 1. Collective Landmark Mapping. (A) We collect smartphone IMU data and free text inputs from users involved in indoor situated tasks. (B) Inertial navigation and large language models identify a small set of landmarks with their position and semantic information from each user’s data. (C) A novel coordinate alignment optimization aggregates multiple user data to generate a unified semantic landmark map. This paper pres… view at source ↗
Figure 2
Figure 2. Data collection app. (A) Trajectory visualization and annotations. (B & C) User￾provided and LLM-generated landmark descriptions. Based on the design goals derived above, we have created Collective Landmark Mapper, which consists of three parts: (i) a mobile application that gathers and visualizes user movement and free text inputs during situated tasks, (ii) a back-end module for identifying landmarks from user-pro… view at source ↗
Figure 3
Figure 3. Coordinate alignment optimization. Multiple record￾ings with different starting loca￾tions and orientations are trans￾formed to be aligned onto a shared coordinate system. Our key technical contribution is a coordinate alignment module that aggregates landmark data provided by multiple users to generate a unified semantic landmark map. Importantly, the coordinate systems used to represent landmark positions are unkn… view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Examples of collected data in the retail restocking checklist tasks: movement trajectories (gray lines), positions of [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: Examples of collected data in the office inspection tasks showing movement trajectories (gray lines), positions of [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Product availability maps generated from restocking checklist tasks. (a-c) Identified landmarks are annotated with [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: Room usage maps generated from office inspection tasks. (a-c) Identified landmarks are annotated with orange circles, [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Quantitative evaluation. We evaluate the positional error (left) between identified and ground-truth landmarks, as [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: Time Analysis. We visualized the histograms for times to complete text entries (left: orange) and those to complete [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: Examples of task progress timelines. The moments of arriving at and leaving from landmarks are marked with black [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]
Figure 11
Figure 11. Figure 11: Simulation experiments: (a) Effect of sampling conditions of landmarks for each data record, (b) effect of positional [PITH_FULL_IMAGE:figures/full_fig_p018_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 33 canonical work pages

  1. [1]

    I am Uncomfortable Sharing What I Can’t See

    Taslima Akter, Bryan Dosono, Tousif Ahmed, Apu Kapadia, and Bryan Semaan. 2020. "I am Uncomfortable Sharing What I Can’t See": Privacy Concerns of the Visually Impaired with Camera Based Assistive Applications. In Proceedings of the USENIX Security Symposium (USENIX Security). 1929–1948

    work page 2020

  2. [2]

    Moustafa Alzantot and Moustafa Youssef. 2012. Crowdinside: Automatic Construction of Indoor Floorplans. In Proceedings of the International Conference on Advances in Geographic Information Systems . 99–108

    work page 2012

  3. [3]

    Isabelle Andersson. 2020. Indoor Positioning Systems in Office Environments . Master’s thesis. Lund University

    work page 2020

  4. [4]

    Mohamed Binalhaj, Hexu Liu, Mohammed Sulaiman, and Osama Abudayyeh. 2021. Mobile Crowdsourcing-based Data Collection for User- Centered Facility Maintenance Management. Canadian Journal of Civil Engineering 48, 12 (2021), 1652–1662. https://doi.org/10.1139/cjce- 2020-0439

    work page doi:10.1139/cjce- 2021

  5. [5]

    André Borrmann, Markus König, Christian Koch, and Jakob Beetz. 2018. Building Information Modeling: Why? What? How? InBuilding Information Modeling: Technology Foundations and Industry Practice , André Borrmann, Markus König, Christian Koch, and Jakob Beetz (Eds.). Springer International Publishing, Cham, 1–24. https://doi.org/10.1007/978-3-319-92862-3_1

    doi:10.1007/978-3-319-92862-3_1 2018

  6. [6]

    Cesar Cadena, Luca Carlone, Henry Carrillo, Yasir Latif, Davide Scaramuzza, José Neira, Ian Reid, and John J. Leonard. 2016. Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age.IEEE Transactions on Robotics (T-RO) 32, 6 (2016), 1309–1332. https://doi.org/10.1109/TRO.2016.2624754

    work page arXiv 2016

  7. [7]

    Lane, Fan Li, Hojung Cha, and Feng Zhao

    Yohan Chon, Nicholas D. Lane, Fan Li, Hojung Cha, and Feng Zhao. 2012. Automatically Characterizing Places with Opportunistic Crowdsensing using Smartphones. In Proceedings of the ACM Conference on Ubiquitous Computing (UbiComp) (UbiComp ’12) . ACM New York, NY, USA, 481–490. https://doi.org/10.1145/2370216.2370288

    work page arXiv 2012

  8. [8]

    Deloitte. 2024. Future of Retail: Profitable Growth through Technology and AI . Corporate Report. Deloitte

    work page 2024

Show all 39 references
  1. [9]

    Ergen, B

    E. Ergen, B. Kula, G. Guven, and D. Artan. 2021. Formalization of Occupant Feedback and Integration with BIM in Office Buildings. Journal of Computing in Civil Engineering 35, 1 (2021), 04020055. https://doi.org/10.1061/(ASCE)CP.1943-5487.0000940

    2021

  2. [10]

    Kapil Garg, Yongsung Kim, Darren Gergle, and Haoqi Zhang. 2019. 4x: A Hybrid Approach for Scaffolding Data Collection and Interest in Low-Effort Participatory Sensing. Proceedings of the ACM on Human-Computer Interaction 3, CSCW (2019), 1–28

    2019

  3. [11]

    Siegwart, and Juan I

    Margarita Grinvald, Fadri Furrer, Tonci Novkovic, Jen Jen Chung, César Cadena, Roland Y. Siegwart, and Juan I. Nieto. 2019. Volumetric Instance-Aware Semantic Mapping and 3D Object Discovery. IEEE Robotics and Automation Letters (RA-L) 4 (2019), 3037–3044. https://api.semantic...

    2019

  4. [12]

    Xiaonan Guo, Eddie CL Chan, Ce Liu, Kaishun Wu, Siyuan Liu, and Lionel M Ni. 2014. Shopprofiler: Profiling Shops with Crowdsourcing Data. In Proceedings of the IEEE Conference on Computer Communications (INFOCOM) . IEEE, 1240–1248

    2014

  5. [13]

    Hayward, K

    S.J. Hayward, K. van Lopik, C. Hinde, and A.A. West. 2022. A Survey of Indoor Location Technologies, Techniques and Applications in Industry. Internet of Things 20 (2022), 100608. https://doi.org/10.1016/j.iot.2022.100608

    2022

  6. [14]

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep Residual Learning for Image Recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . 770–778

    2016

  7. [15]

    Sachini Herath, David Caruso, Chen Liu, Yufan Chen, and Yasutaka Furukawa. 2022. Neural inertial localization. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . IEEE

    2022

  8. [16]

    Sachini Herath, Saghar Irandoust, Bowen Chen, Yiming Qian, Pyojin Kim, and Yasutaka Furukawa. 2021. Fusion-dhl: Wifi, imu, and floorplan fusion for dense history of locations in indoor environments. In Proceedings of the IEEE International Conference on Robotics and Automation...

    2021

  9. [17]

    Xinyu Hou and Jeroen Bergmann. 2020. Pedestrian dead reckoning with wearable sensors: A systematic review. IEEE Sensors Journal 21, 1 (2020), 143–152

    2020

  10. [18]

    Haosheng Huang, Georg Gartner, Jukka M Krisp, Martin Raubal, and Nico Van de Weghe. 2018. Location based Services: Ongoing Evolution and Research Agenda. Journal of Location Based Services 12, 2 (2018), 63–93

    2018

  11. [19]

    Yongsung Kim, Darren Gergle, and Haoqi Zhang. 2018. Hit-or-Wait: Coordinating Opportunistic Low-effort Contributions to Achieve Global Outcomes in On-the-go Crowdsourcing. In Proceedings of the CHI Conference on Human Factors in Computing Systems (CHI) . Proc. ACM Interact. Mo...

    2018

  12. [20]

    Daniel Laidig and Thomas Seel. 2023. VQF: Highly Accurate IMU Orientation Estimation with Bias Estimation and Magnetic Disturbance Rejection. Information Fusion 91 (2023), 187–204

    2023

  13. [21]

    Pierre-Yves Lajoie, Benjamin Ramtoula, Fang Wu, and Giovanni Beltrame. 2022. Towards collaborative simultaneous localization and mapping: a survey of the current research landscape. Field Robotics 2 (2022), 971–1000

    2022

  14. [22]

    Chao Li, Wennan Chai, Xiaohui Yang, and Qingdang Li. 2022. Crowdsourcing-Based Indoor Semantic Map Construction and Localization Using Graph Optimization. Sensors 22, 16 (2022), 6263

    2022

  15. [23]

    Katherine Lin, Henry Spindell, Scott Cambo, Yongsung Kim, and Haoqi Zhang. 2016. Habitsourcing: Sensing the Environment through Immersive, Habit-Building Experiences. In Proceedings of the Annual Symposium on User Interface Software and Technology (UIST) . ACM New York, NY, US...

    2016

  16. [24]

    Wenxin Liu, David Caruso, Eddy Ilg, Jing Dong, Anastasios I Mourikis, Kostas Daniilidis, Vijay Kumar, and Jakob Engel. 2020. Tlio: Tight learned inertial odometry. IEEE Robotics and Automation Letters (RA-L) 5, 4 (2020), 5653–5660

    2020

  17. [25]

    Roie Melamed. 2016. Indoor Localization: Challenges and Opportunities. In Proceedings of the International Conference on Mobile Software Engineering and Systems (MOBILESoft) . ACM New York, NY, USA, 1–2. https://doi.org/10.1145/2897073.2897074

    2016

  18. [26]

    Muhammad Naeem, Waleed Ejaz, Muhammad Iqbal, Farkhund Iqbal, Alagan Anpalagan, and Joel JPC Rodrigues. 2020. Efficient Scheduling of Video Camera Sensor Networks for IoT Systems in Smart Cities. Transactions on Emerging Telecommunications Technologies 31, 5 (2020), e3798

    2020

  19. [27]

    Petteri Nurmi, Antti Salovaara, Sourav Bhattacharya, Teemu Pulkkinen, and Gerrit Kahl. 2011. Influence of Landmark-based Navigation Instructions on User Attention in Indoor Smart Spaces. In Proceedings of the International Conference on Intelligent User Interfaces (IUI) . ACM ...

    2011

  20. [28]

    Sangkeun Park, Sujin Kwon, and Uichin Lee. 2018. Campuswatch: Exploring Communitysourced Patrolling with Pervasive Mobile Technology. Proceedings of the ACM on Human-Computer Interaction 2, CSCW (2018), 1–25

    2018

  21. [29]

    Sonia Raychaudhuri and Angel X Chang. 2025. Semantic Mapping in Indoor Embodied AI–A Comprehensive Survey and Future Directions. arXiv preprint arXiv:2501.05750 (2025)

    2025 arXiv

  22. [30]

    Antoni Rosinol, Marcus Abate, Yun Chang, and Luca Carlone. 2019. Kimera: an Open-Source Library for Real-Time Metric-Semantic Localization and Mapping. In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) . 1689–1696. https: //api.semanticschol...

    2019

  23. [31]

    Swapnil Sayan Saha, Sandeep Singh Sandha, Luis Antonio Garcia, and Mani Srivastava. 2022. TinyOdom: Hardware-aware efficient neural inertial navigation. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT) 6, 2, Article 71 (2022)

    2022

  24. [32]

    Smitha Sheshadri and Kotaro Hara. 2024. Conversational Localization: Indoor Human Localization through Intelligent Conversation. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT) 7, 4 (2024), 1–32

    2024

  25. [33]

    Scott Sun, Dennis Melamed, and Kris Kitani. 2021. Idol: Inertial deep orientation-estimation and localization. In Proceedngs of the AAAI Conference on Artificial Intelligence (AAAI). AAAI Press

    2021

  26. [34]

    Xiaoqiang Teng, Deke Guo, Yulan Guo, Xiang Zhao, and Zhong Liu. 2018. SISE: Self-Updating of Indoor Semantic Floorplans for General Entities. IEEE Transactions on Mobile Computing 17, 11 (2018), 2646–2659

    2018

  27. [35]

    Oliver J Woodman. 2007. An introduction to inertial navigation . Technical Report. University of Cambridge, Computer Laboratory

    2007

  28. [36]

    Yuan Wu, Hai-Bing Zhu, Qing-Xiu Du, and Shu-Ming Tang. 2019. A survey of the research status of pedestrian dead reckoning systems based on inertial sensors. International Journal of Automation and Computing 16 (2019), 65–83

    2019

  29. [37]

    Hang Yan, Sachini Herath, and Yasutaka Furukawa. 2020. RoNIN: Robust Neural Inertial Navigation in the Wild: Benchmark, Evaluations, & New Methods. In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) . IEEE

    2020

  30. [38]

    Baoding Zhou, Wei Ma, Qingquan Li, Naser El-Sheimy, Qingzhou Mao, You Li, Fuqiang Gu, Lian Huang, and Jiasong Zhu. 2021. Crowdsourcing-based Indoor Mapping using Smartphones: A Survey. ISPRS Journal of Photogrammetry and Remote Sensing 177 (2021), 131–146. https://doi.org/10.1...

    2021 doi

  31. [39]

    Heng Zhou and Takuya Maekawa. 2023. GPS-assisted Indoor Pedestrian Dead Reckoning. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT) 6, 4 (2023), 1–36. Received 1 May 2025; revised 1 May 2025; accepted 1 May 2025 Proc. ACM Interact. Mo...

    2023

Pith tools

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