Pith. sign in

REVIEW 5 major objections 5 minor 38 references

Permutation-Invariant Transformer Neural Architectures for Set-Based Indoor Localization Using Learned RSSI Embeddings

T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A Set Transformer that treats Wi-Fi scans as unordered sets of (BSSID, RSSI) pairs positions users indoors with competitive accuracy, ranking second behind a plain LSTM in all three experiments.

desk verdict A modest, honest benchmark paper that will live or die on the quality of its unvalidated ground-truth labels; the reported LSTM-vs-SetTransformer gaps are small enough that even a meter of label noise could change the ranking. read the letter →

arxiv 2506.00656 v1 pith:QWJGUQQB submitted 2025-05-31 cs.LG cs.AI

Aris J. Aristorenas This is my paper
classification cs.LGcs.AI
keywords indoorlocalizationRSSIpermutationinvarianceSetTransformerlearnedembeddingsBSSIDWi-Fifingerprintingset-basedneuralnetworks
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

Indoor positioning often relies on Wi-Fi signal strengths, but each scan arrives with a variable number of access points in arbitrary order, a structure many neural models force into fixed vectors. This paper tests whether a Set Transformer, an attention architecture designed for unordered sets, can map raw scans of (BSSID, RSSI) pairs directly to coordinates. On a six-building campus dataset it can: the Set Transformer ranked second behind a simple LSTM in every experiment, with mean errors of 3.82 m, 6.30 m, and 3.53 m, and it preserved building and floor structure in its predicted paths. This is evidence that permutation-invariant set models are a principled, ordering-free option for signal-based localization, even though the plain LSTM remains the strongest baseline on this data.

What carries the argument

The core mechanism is the Set Transformer, which uses Set Attention Blocks (multi-head self-attention with residual connections and layer normalization) followed by Pooling by Multihead Attention, in which a learned seed vector attends over the whole set to produce a fixed-size summary. Each BSSID is first mapped to a learned embedding vector and concatenated with its RSSI value, so the model reasons over access-point relationships rather than raw MAC identifiers. That design delivers permutation invariance, accepts scans of any length without padding, and assigns random initial embeddings to BSSIDs not seen during training.

What would settle it

Survey a subset of the test paths with a high-precision reference (for example, a laser rangefinder or total station) and recompute the mean errors in Table 2; systematic label offsets larger than roughly one meter would mean the reported ranking is not settled.

Watch

Extended reading notes

Core claim

The paper's central claim is that permutation-invariant set processing is a natural inductive bias for RSSI-based indoor localization. Treating each scan as an unordered set, mapping each BSSID to a learned dense embedding, and aggregating the set with Set Transformer attention blocks yields accurate predictions in single-building, multi-building, and multi-floor settings. Across the three tasks the Set Transformer was the second-best model in every case, with mean errors of 3.82 m, 6.30 m, and 3.53 m, behind an LSTM that reads the same pairs sorted by signal strength (2.23 m, 3.13 m, and 2.44 m). The paper presents this as evidence that set-based architectures provide accuracy alongside architectural generality, including robustness to arbitrary ordering, missing access points, and previously unseen BSSIDs.

Load-bearing premise

The ground-truth coordinates, produced by post-hoc satellite mapping and manual annotation of walking paths, are accurate at the meter scale the paper reports; if those labels contain systematic drift, every error number in Table 2 and the ranking between the LSTM and the Set Transformer becomes unreliable.

Editorial extensions

If this is right

  • Indoor localization can proceed without imposing a canonical ordering on access-point detections: the Set Transformer consumes scans directly as sets and still recovers corridor-level spatial structure.
  • Newly seen access points need not trigger retraining; randomly initialized embeddings for unseen BSSIDs let the model continue making sensible predictions at inference time.
  • Set-based attention generalizes across physically distinct buildings and floors, keeping floor predictions separated in 3D where MLPs and vanilla RNNs leak across floors.
  • A simple LSTM remains a powerful baseline for RSSI localization even when the input is not inherently temporal, so future set-model comparisons should include recurrent models.
  • The reported standard deviations keep the Set Transformer's errors in the same general range as the LSTM's, so the second-place ranking reflects broadly consistent accuracy rather than isolated outliers.

Reading between the lines

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

  • Shuffling the access-point order at test time would give a direct, cheap check of the paper's central motivation: the Set Transformer's predictions should be unchanged by construction, while the LSTM's may shift if it has latched onto ordering cues; a large LSTM degradation under permutation would turn the set model's robustness advantage from hypothesis into measurement.
  • Because ground truth came from post-hoc satellite mapping, resurveying a subset of paths with higher-precision equipment would show whether the roughly 1.5 to 3 meter gaps between the Set Transformer and the LSTM are stable or within label error.
  • The dataset contains only straight-line hallway walks in university buildings, so the multi-floor and multi-building conclusions may not transfer to open-plan or radio-noisy environments; testing in a mall, airport, or hospital would be the natural next stress test.
  • The paper leaves batching and attention masking as future work; scaling the Set Transformer past batch size one is a plausible route to closing the gap with the LSTM, though the paper itself does not claim that gain.
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

5 major / 5 minor

Summary. The paper proposes a permutation-invariant neural architecture for RSSI-based indoor localization. Each Wi-Fi scan is treated as an unordered set of (BSSID, RSSI) pairs, with BSSIDs mapped to learned embeddings that are concatenated with the RSSI value and processed by a Set Transformer. The method is evaluated on a self-collected campus dataset spanning six buildings on three tasks: single-building/single-floor (E1), multiple-building/first-floor (E2), and single-building/multiple-floor (E3). The authors compare against MLP, RNN, LSTM, and attention baselines, reporting that the LSTM achieves the lowest mean localization error in all three experiments (2.23 m, 3.13 m, 2.44 m) while the Set Transformer ranks second (3.82 m, 6.30 m, 3.53 m). The paper claims that the Set Transformer is a competitive permutation-invariant alternative and that learned BSSID embeddings enable generalization to unseen BSSIDs.

Significance. If the reported rankings are reliable, the paper provides a useful empirical datapoint: a permutation-invariant set-based model can handle sparse, unordered RSSI scans and perform competitively, though a simple LSTM remains stronger on this dataset. The problem formulation, with explicit permutation invariance and variable-cardinality handling, is clear, and the paper is honest in reporting that the proposed model does not beat the LSTM. The main value is the comparison itself and the application of Set Transformers to a new task. However, the paper's broader claims about cross-domain generalization and unseen-BSSID robustness are not supported by the experimental design, and the ground-truth labeling procedure is not validated at the meter scale that the conclusions depend on. The paper currently lacks the evidence needed to establish those generalization claims, though the core ranking could survive additional validation.

major comments (5)
  1. [Section 5.1, Table 2] The central empirical ranking rests on ground-truth coordinates obtained by post-hoc satellite mapping with manual annotation, yet the paper provides no quantitative validation of these labels. No control-point survey, inter-annotator comparison, or registration error estimate is reported. This is load-bearing because the reported differences between LSTM and Set Transformer are about 1.1 to 3.2 m (E1: 2.23 vs 3.82; E2: 3.13 vs 6.30; E3: 2.44 vs 3.53). If label noise or systematic drift is on the order of 1–2 m, the ranking in Table 2 could change. Section 5.3 and 5.4 describe integrity checks, but none of them validates the geometric accuracy of the labels. Please add a validation study of the annotation procedure or, failing that, temper the accuracy claims accordingly.
  2. [Abstract, Section 6.2] The abstract says the model 'maintains performance across physically distinct domains,' and Section 8.2 claims generalization across buildings and floors, but E2 and E3 are not cross-domain generalization experiments. Section 6.2 states that the training and test sets 'assume the same distribution' and that the held-out test set 'reasonably contains examples from all the buildings' (E2) and 'from each floor' (E3). Thus the test sets are sampled from the same buildings and floors seen in training, so the tasks measure interpolation within familiar domains, not generalization to unseen domains. This overclaim should be corrected, or the experiments should be redesigned to hold out entire buildings or floors.
  3. [Sections 4.1/4.2 vs Section 6.1] There is a direct contradiction about batch size. Section 4.1 and Section 4.2 state that each RSSI set is processed individually with batch size = 1 to avoid padding, while Section 6.1 says all models were trained with an identical batch size of 32. This inconsistency makes the experimental setup unclear and affects reproducibility. Please specify the actual batch size used for the Set Transformer and explain how variable-length sets are handled if the batch size is greater than one.
  4. [Sections 4.9 and 8.2] The claim that the model generalizes to unseen BSSIDs via randomly initialized embeddings is not tested in isolation. Section 4.9 says many BSSIDs in D_test 'may not be present in any training set,' but because E2 and E3 test sets contain examples from all buildings and floors present in training, it is unclear whether any test BSSIDs are truly unseen. Section 8.2 then states that the Set Transformer 'maintains competitive accuracy' on unseen BSSIDs, which is not demonstrated by the reported experiments. To support this claim, an experiment should hold out a set of BSSIDs during training and evaluate on scans containing only those BSSIDs, or the claim should be removed.
  5. [Sections 4.7 and 6.1] The Set Transformer architecture is under-specified, so the 'matched hyperparameters and capacity constraints' assertion in Section 6.1 cannot be verified. The paper does not report the number of Set Attention Blocks, the number of attention heads, the embedding dimension d, the hidden dimensionality, or the total parameter count. Without this information, the reader cannot judge whether the comparison to the LSTM is fair in terms of model capacity, nor can the experiments be reproduced. Please include a full architecture specification and, ideally, parameter counts for all models.
minor comments (5)
  1. [Figure 1] The text repeatedly refers to Figure 1 for qualitative assessment, but the figure itself is not included in the manuscript; only a caption is present. Please include the figure or remove the references until the figure is available.
  2. [Section 5.5] The sentence 'Domain floor plans were also acquired to aid in cross-verification of path alignments' (Section 5.5) promises a verification step, but no results of that verification are reported. Either describe what the cross-verification showed or mention it as a limitation.
  3. [Section 6.1] The list of baselines in Section 6.1 mentions four baselines plus the Set Transformer, but the text says '4 baseline models' in Section 7; please check the count for consistency.
  4. [Section 4.9] The multi-task extension with the auxiliary classification loss (Eq. 16) is described but never evaluated; the paper does not report results for the multi-task variant. Either present those results or clearly state that the multi-task model is not part of the main evaluation.
  5. [Section 8.5] The limitations section is candid, but it does not mention the lack of statistical significance testing. Table 2 reports per-sample mean and standard deviation, but there is no assessment of run-to-run variance (e.g., multiple seeds) or significance of the differences between LSTM and Set Transformer. A brief note would help calibrate the strength of the ranking.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the paper is an empirical benchmark whose ranking claims rest on held-out test evaluation, not on self-referential or fitted definitions.

full rationale

The paper's central claims are empirical: a Set Transformer with learned BSSID embeddings is compared against MLP, RNN, LSTM, and attention baselines on three supervised localization tasks, with error reported on held-out test scans (Eq. 17, Table 2). No quantity claimed as a prediction is defined in terms of the outcome it predicts. The loss (Eq. 15) is a standard supervised MSE between model outputs and normalized ground-truth coordinates; the ground-truth coordinates come from post-hoc satellite-based mapping described in Section 5.1, which is an external labeling procedure, not a fitted parameter of the model. The Set Transformer is adopted from an external publication (Lee et al. [25]) rather than from the authors' prior work, so no self-citation chain carries the argument. The claim about handling 'previously unseen BSSIDs' (Section 3, Section 8.2) is a stated capability supported by assigning randomly initialized embeddings at inference time; it is not tested in isolation, but that is a missing measurement rather than a circular reduction. The acknowledged limitations in Section 8.5 (straight-line collection paths, homogeneous buildings, absence of ablations, no latency benchmarks) concern generalizability and completeness, not circularity. The skeptical concern about unvalidated ground-truth label accuracy is a data-quality and measurement-validity risk, not evidence that any derivation step reduces to its own input. Overall, the derivation chain is self-contained with respect to circularity: every numerical result is an empirical observation on held-out data, and no fitted input is renamed as a prediction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new theoretical entities; its contributions are empirical. Its central results rest on unexamined data-quality assumptions (label accuracy, RSSI informativeness) and unreported architectural hyperparameters.

free parameters (3)
  • BSSID embedding dimension d
    The size of the learned embedding table is a hand-chosen capacity parameter that controls how much identity information the model can memorize; it is never reported.
  • Set Transformer depth and heads
    Number of Set Attention Blocks, attention heads, and seed vectors are not specified, though the paper claims matched capacity across models.
  • Multi-task loss weight lambda = 1
    Section 4.9 sets the classification/regression weighting to 1 without tuning or sensitivity analysis.
assumptions (4)
  • domain assumption Ground-truth UTM coordinates from post-hoc satellite mapping are sufficiently accurate for meter-level localization labels.
    Section 5.1 describes manual annotation using satellite mapping software; the entire supervised training and error reporting depends on this assumption.
  • domain assumption RSSI from detected Wi-Fi access points carries enough spatial information to regress 2D positions.
    This is the underlying premise of RSSI fingerprinting and is not tested independently in the paper.
  • standard math Set Transformer is a valid permutation-invariant set function approximator.
    The paper relies on the universality results of Lee et al. (2019) without proving them.
  • domain assumption Training and test sets in E2 and E3 share the same building and floor distribution.
    Section 6.2 states test sets reasonably contain examples from all buildings and floors; the paper calls this generalization, but it is not domain shift.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Permutation-Invariant Transformer Neural Architectures for Set-Based Indoor Localization Using Learned RSSI Embeddings." pith.science (2026). https://pith.science/paper/QWJGUQQB

@misc{pith2026250600656,
  author       = {Pith},
  title        = {Pith review of: Permutation-Invariant Transformer Neural Architectures for Set-Based Indoor Localization Using Learned RSSI Embeddings},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QWJGUQQB}},
  note         = {Machine review of arXiv:2506.00656}
}
read the original abstract

We propose a permutation-invariant neural architecture for indoor localization using RSSI scans from Wi-Fi access points. Each scan is modeled as an unordered set of (BSSID, RSSI) pairs, where BSSIDs are mapped to learned embeddings and concatenated with signal strength. These are processed by a Set Transformer, enabling the model to handle variable-length, sparse inputs while learning attention-based representations over access point relationships. We evaluate the model on a dataset collected across a campus environment consisting of six buildings. Results show that the model accurately recovers fine-grained spatial structure and maintains performance across physically distinct domains. In our experiments, a simple LSTM consistently outperformed all other models, achieving the lowest mean localization error across three tasks (E1 - E3), with average errors as low as 2.23 m. The Set Transformer performed competitively, ranking second in every experiment and outperforming the MLP, RNN, and basic attention models, particularly in scenarios involving multiple buildings (E2) and multiple floors (E3). Performance degraded most in E2, where signal conditions varied substantially across buildings, highlighting the importance of architectural robustness to domain diversity. This work demonstrates that set-based neural models are a natural fit for signal-based localization, offering a principled approach to handling sparse, unordered inputs in real-world positioning tasks.

Figures

Figures reproduced from arXiv: 2506.00656 by the authors.

Figure 1
Figure 1. Comparison of model predictions across experiments (columns: E1, E2, E3) and model [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

38 extracted references · 30 canonical work pages

  1. [1]

    A bluetooth location- based indoor positioning system for asset tracking in warehouse

    Carman Ka Man Lee, CM Ip, Taezoon Park, and Sui Ying Chung. A bluetooth location- based indoor positioning system for asset tracking in warehouse. In 2019 IEEE international conference on industrial engineering and engineering management (IEEM), pages 1408–1412. IEEE, 2019

    work page 2019

  2. [2]

    A hybrid indoor positioning for asset tracking using bluetooth low energy and wi-fi

    Chun-Hao Kao, Rong-Shue Hsiao, Tian-Xiang Chen, Po-Shao Chen, and Mei-Jin Pan. A hybrid indoor positioning for asset tracking using bluetooth low energy and wi-fi. In 2017 IEEE international conference on consumer electronics-Taiwan (ICCE-TW) , pages 63–64. IEEE, 2017

    work page 2017

  3. [3]

    A survey of indoor location technologies, techniques and applications in industry

    SJ Hayward, Katherine van Lopik, Christopher Hinde, and Andrew A West. A survey of indoor location technologies, techniques and applications in industry. Internet of Things, 20:100608, 2022

    work page 2022

  4. [4]

    Indoor positioning systems of mobile robots: A review

    Jiahao Huang, Steffen Junginger, Hui Liu, and Kerstin Thurow. Indoor positioning systems of mobile robots: A review. Robotics, 12(2):47, 2023

    work page 2023

  5. [5]

    In- door positioning of a robotic walking assistant for large public environments.IEEE Transactions on Instrumentation and Measurement, 64(11):2965–2976, 2015

    Payam Nazemzadeh, Federico Moro, Daniele Fontanelli, David Macii, and Luigi Palopoli. In- door positioning of a robotic walking assistant for large public environments.IEEE Transactions on Instrumentation and Measurement, 64(11):2965–2976, 2015

    work page 2015

  6. [6]

    The study on using passive rfid tags for indoor positioning

    SL Ting, SK Kwok, Albert HC Tsang, and George TS Ho. The study on using passive rfid tags for indoor positioning. International journal of engineering business management, 3:8, 2011. 14

    work page 2011

  7. [7]

    Review of rfid-based indoor positioning technology

    Jingkai Zhu and He Xu. Review of rfid-based indoor positioning technology. In Innovative Mobile and Internet Services in Ubiquitous Computing: Proceedings of the 12th International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing (IMIS-2018), pages 632–641. Springer, 2019

    work page 2018

  8. [8]

    Rssi based bluetooth low energy indoor positioning

    Zhu Jianyong, Luo Haiyong, Chen Zili, and Li Zhaohui. Rssi based bluetooth low energy indoor positioning. In 2014 International Conference on Indoor Positioning and Indoor Navigation (IPIN), pages 526–533. IEEE, 2014

    work page 2014

Show all 38 references
  1. [9]

    A low cost indoor positioning system using bluetooth low energy

    Lu Bai, Fabio Ciravegna, Raymond Bond, and Maurice Mulvenna. A low cost indoor positioning system using bluetooth low energy. Ieee Access, 8:136858–136871, 2020

    2020

  2. [10]

    Ultra wideband indoor positioning technologies: Analysis and recent advances

    Abdulrahman Alarifi, AbdulMalik Al-Salman, Mansour Alsaleh, Ahmad Alnafessah, Suheer Al-Hadhrami, Mai A Al-Ammar, and Hend S Al-Khalifa. Ultra wideband indoor positioning technologies: Analysis and recent advances. Sensors, 16(5):707, 2016

    2016

  3. [11]

    Precise indoor position- ing using uwb: A review of methods, algorithms and implementations

    Fazeelat Mazhar, Muhammad Gufran Khan, and Benny Sällberg. Precise indoor position- ing using uwb: A review of methods, algorithms and implementations. Wireless Personal Communications, 97(3):4467–4491, 2017

    2017

  4. [12]

    Deep learning methods for fingerprint-based indoor positioning: A review

    Fahad Alhomayani and Mohammad H Mahoor. Deep learning methods for fingerprint-based indoor positioning: A review. Journal of Location Based Services, 14(3):129–200, 2020

    2020

  5. [13]

    Intelligent indoor positioning based on artificial neural networks

    Wen-Long Chin, Cheng-Che Hsieh, David Shiung, and Tao Jiang. Intelligent indoor positioning based on artificial neural networks. IEEE Network, 34(6):164–170, 2020

    2020

  6. [14]

    Ccpos: Wifi fingerprint indoor positioning system based on cdae-cnn

    Feng Qin, Tao Zuo, and Xing Wang. Ccpos: Wifi fingerprint indoor positioning system based on cdae-cnn. Sensors, 21(4):1114, 2021

    2021

  7. [15]

    Improved cnn-based magnetic indoor positioning system using attention mechanism

    Mahdi Abid, Paul Compagnon, and Grégoire Lefebvre. Improved cnn-based magnetic indoor positioning system using attention mechanism. In 2021 International Conference on Indoor Positioning and Indoor Navigation (IPIN), pages 1–8. IEEE, 2021

    2021

  8. [16]

    A fast indoor positioning using a knowledge-distilled convolutional neural network (kd-cnn)

    Aqilah Binti Mazlan, Yin Hoe Ng, and Chee Keong Tan. A fast indoor positioning using a knowledge-distilled convolutional neural network (kd-cnn). IEEE access, 10:65326–65338, 2022

    2022

  9. [17]

    Cnn based indoor localization using rss time-series

    Mai Ibrahim, Marwan Torki, and Mustafa ElNainay. Cnn based indoor localization using rss time-series. In 2018 IEEE symposium on computers and communications (ISCC), pages 01044–01049. IEEE, 2018

    2018

  10. [18]

    Indoor positioning based on fingerprint-image and deep learning

    Wenhua Shao, Haiyong Luo, Fang Zhao, Yan Ma, Zhongliang Zhao, and Antonino Crivello. Indoor positioning based on fingerprint-image and deep learning. Ieee Access, 6:74699–74712, 2018

    2018

  11. [19]

    Recurrent neural networks model for wifi- based indoor positioning system

    Yuan Lukito and Antonius Rachmat Chrismanto. Recurrent neural networks model for wifi- based indoor positioning system. In 2017 International Conference on Smart Cities, Automation & Intelligent Computing Systems (ICON-SONICS), pages 121–125. IEEE, 2017

    2017

  12. [20]

    Dl-rnn: An accurate indoor localization method via double rnns

    Siqi Bai, Mingjiang Yan, Qun Wan, Long He, Xinrui Wang, and Junlin Li. Dl-rnn: An accurate indoor localization method via double rnns. IEEE Sensors Journal, 20(1):286–295, 2019

    2019

  13. [21]

    Recurrent neural networks for accurate rssi indoor localization

    Minh Tu Hoang, Brosnan Yuen, Xiaodai Dong, Tao Lu, Robert Westendorp, and Kishore Reddy. Recurrent neural networks for accurate rssi indoor localization. IEEE Internet of Things Journal, 6(6):10639–10651, 2019

    2019

  14. [22]

    An lstm-based indoor positioning method using wi-fi signals

    Ayesha Sahar and Dongsoo Han. An lstm-based indoor positioning method using wi-fi signals. In Proceedings of the 2nd International Conference on Vision, Image and Signal Processing, pages 1–5, 2018

    2018

  15. [23]

    Sdr-fi: Deep- learning-based indoor positioning via software-defined radio

    Erick Schmidt, Devasena Inupakutika, Rahul Mundlamuri, and David Akopian. Sdr-fi: Deep- learning-based indoor positioning via software-defined radio. IEEE Access, 7:145784–145797, 2019. 15

    2019

  16. [24]

    Deep learning based wireless localization for indoor navigation

    Roshan Ayyalasomayajula, Aditya Arun, Chenfeng Wu, Sanatan Sharma, Abhishek Rajkumar Sethi, Deepak Vasisht, and Dinesh Bharadia. Deep learning based wireless localization for indoor navigation. In Proceedings of the 26th Annual International Conference on Mobile Computing and ...

    2020

  17. [25]

    Set transformer: A framework for attention-based permutation-invariant neural networks

    Juho Lee, Yoonho Lee, Jungtaek Kim, Adam Kosiorek, Seungjin Choi, and Yee Whye Teh. Set transformer: A framework for attention-based permutation-invariant neural networks. In International conference on machine learning, pages 3744–3753. PMLR, 2019

    2019

  18. [26]

    Deep sets

    Manzil Zaheer, Satwik Kottur, Siamak Ravanbakhsh, Barnabas Poczos, Russ R Salakhutdinov, and Alexander J Smola. Deep sets. Advances in neural information processing systems, 30, 2017

    2017

  19. [27]

    Order matters: Sequence to sequence for sets

    Oriol Vinyals, Samy Bengio, and Manjunath Kudlur. Order matters: Sequence to sequence for sets. arXiv preprint arXiv:1511.06391, 2015

    2015 arXiv

  20. [28]

    Perceiver: General perception with iterative attention

    Andrew Jaegle, Felix Gimeno, Andy Brock, Oriol Vinyals, Andrew Zisserman, and Joao Carreira. Perceiver: General perception with iterative attention. In International conference on machine learning, pages 4651–4664. PMLR, 2021

    2021

  21. [29]

    Janossy pooling: Learning deep permutation-invariant functions for variable-size inputs

    Ryan L Murphy, Balasubramaniam Srinivasan, Vinayak Rao, and Bruno Ribeiro. Janossy pooling: Learning deep permutation-invariant functions for variable-size inputs. arXiv preprint arXiv:1811.01900, 2018

    2018 arXiv

  22. [30]

    Setvae: Learning hierarchical composition for generative modeling of set-structured data

    Jinwoo Kim, Jaehoon Yoo, Juho Lee, and Seunghoon Hong. Setvae: Learning hierarchical composition for generative modeling of set-structured data. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15059–15068, 2021

    2021

  23. [31]

    Object-centric learning with slot attention

    Francesco Locatello, Dirk Weissenborn, Thomas Unterthiner, Aravindh Mahendran, Georg Heigold, Jakob Uszkoreit, Alexey Dosovitskiy, and Thomas Kipf. Object-centric learning with slot attention. Advances in neural information processing systems, 33:11525–11538, 2020

    2020

  24. [32]

    Fspool: Learning set representations with featurewise sort pooling

    Yan Zhang, Jonathon Hare, and Adam Prügel-Bennett. Fspool: Learning set representations with featurewise sort pooling. arXiv preprint arXiv:1906.02795, 2019

    1906 arXiv

  25. [33]

    Setformer is what you need for vision and language

    Pourya Shamsolmoali, Masoumeh Zareapoor, Eric Granger, and Michael Felsberg. Setformer is what you need for vision and language. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 4713–4721, 2024

    2024

  26. [34]

    Deepfi: Deep learning for indoor fingerprinting using channel state information

    Xuyu Wang, Lingjun Gao, Shiwen Mao, and Santosh Pandey. Deepfi: Deep learning for indoor fingerprinting using channel state information. In 2015 IEEE wireless communications and networking conference (WCNC), pages 1666–1671. IEEE, 2015

    2015

  27. [35]

    Cnnloc: Deep-learning based indoor localization with wifi fingerprinting

    Xudong Song, Xiaochen Fan, Xiangjian He, Chaocan Xiang, Qianwen Ye, Xiang Huang, Gengfa Fang, Liming Luke Chen, Jing Qin, and Zumin Wang. Cnnloc: Deep-learning based indoor localization with wifi fingerprinting. In 2019 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Adv...

    2019

  28. [36]

    Semi-sequential probabilistic model for indoor localization enhancement

    Minh Tu Hoang, Brosnan Yuen, Xiaodai Dong, Tao Lu, Robert Westendorp, and Kishore Reddy Tarimala. Semi-sequential probabilistic model for indoor localization enhancement. IEEE Sensors Journal, 20(11):6160–6169, 2020

    2020

  29. [37]

    Exploring indoor localization with transformer-based models: a cnn-transformer hybrid approach for wifi fingerprinting

    Nicu Savin. Exploring indoor localization with transformer-based models: a cnn-transformer hybrid approach for wifi fingerprinting. B.S. thesis, University of Twente, 2023

    2023

  30. [38]

    Hytra: Hyperclass transformer for wifi fingerprinting-based indoor localization

    Muneeb Nasir, Kiara Esguerra, Ibrahima Faye, Tong Boon Tang, Mazlaini Yahya, Afidalina Tumian, and Eric Tatt Wei Ho. Hytra: Hyperclass transformer for wifi fingerprinting-based indoor localization. Transactions on Energy Systems and Engineering Applications, 5(1):1–24, 2024. 16

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.