Plant Bioelectric Early Warning Systems: A Five-Year Investigation into Human-Plant Electromagnetic Communication

Peter A. Gloor

REVIEW 4 major objections 6 minor 16 references

Plant Bioelectric Early Warning Systems: A Five-Year Investigation into Human-Plant Electromagnetic Communication

T0 review · 4 major / 6 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 five-year study claims basil plants generate voltage signals that let a deep-learning classifier read a nearby human's emotional state at 97% accuracy, with shuffled-label controls at 30%.

desk verdict The paper's own Table 1 disproves its headline 97% seven-way emotion decoding, and the rest is a re-analysis of the authors' earlier work with a speculative wrapper. read the letter →

arxiv 2506.04132 v1 pith:UG2HOG22 submitted 2025-06-04 q-bio.OT cs.AI

Peter A. Gloor This is my paper

classification q-bio.OTcs.AI

keywords plantelectrophysiologybioelectricsignalingemotionrecognitionmachinelearninghuman-plantinteractionearlywarningsystemsOcimumbasilicummel-spectrogramclassification

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 basil plants produce bioelectric voltage patterns that carry decodable information about a nearby human's emotional state, and that this reflects an evolved early-warning system against herbivores. Using electrode recordings transcribed into mel-spectrograms, a transfer-learning neural network reportedly reached 97% accuracy on seven emotion categories, while shuffled labels yielded 30% accuracy, which the authors take as evidence the signal is genuine. The paper's own per-class table shows that the 97% figure is driven by the three dominant classes (neutral, sad, happy); the rare classes (disgusted, fearful, surprised) are never recognized, and the macro F1 is 0.56. The authors integrate five years of related experiments—individual recognition, gesture detection, voice response, stress prediction, and sleep staging—into the hypothesis that high-palatability plant species evolved pre-contact detection of approaching animals.

What carries the argument

The machinery is a transfer-learning classifier on mel-spectrogram images of plant voltage. A differential amplifier (ESP32 with INA128) records leaf-to-soil voltage at 400 Hz; the signal is bandpass filtered at 0.1–50 Hz, segmented into 20-second windows, and converted into 64-bin mel-spectrograms. A ResNet50 network pretrained on a large general image corpus reuses its visual feature extractors to classify these spectrogram images into seven emotion classes. The mel-spectrogram is the pivotal object: it turns a one-dimensional voltage trace into a two-dimensional image whose time-frequency structure is what the network learns to associate with emotions.

What would settle it

A decisive control would place the participant in an electrically shielded, sound-isolated room with no sightline, no shared air, and no audio or video link to the plant while the same electrode rig records. If classification accuracy stays near 97%, emotion-specific biophysical coupling is supported; if accuracy falls to chance, the classifier was reading concurrent human activity rather than emotional state.

Watch

Extended reading notes

Core claim

The central claim is that Ocimum basilicum plants generate bioelectric potential fluctuations that differ systematically with the emotional state of a nearby human, and that these differences are large enough for a machine-learning classifier to read them. The paper reports 97% overall accuracy on a held-out set of 898 spectrogram windows, versus 30% for shuffled labels, and interprets this contrast as evidence that plant voltage patterns contain genuine information about internal human state. It generalizes this into an early-warning hypothesis: plants under strong herbivore pressure evolved the ability to detect approaching animals through bioelectric or other pre-contact cues, allowing defense mobilization before damage occurs. The reported accuracy is not uniform across the seven classes: neutral, sad, and happy dominate the test set and are classified well, while disgusted, fearful, and surprised are never correctly classified, producing a macro F1 of 0.56.

Load-bearing premise

The plant-voltage differences during emotion sessions are caused by the participant's internal emotional state rather than by correlated body movement, breathing, voice, temperature changes, or electrode drift.

Editorial extensions

If this is right

If the 97% result replicates, plant bioelectric monitoring could offer a non-invasive, continuous window onto human emotional state that works at a distance.
The early-warning hypothesis predicts that high-palatability species (basil, lettuce, tomato) respond most strongly while defended species (zucchini, corn, orchids) respond weakly; the paper reports exactly this species pattern.
The same pipeline is claimed to extend to stress prediction above 90% accuracy and to sleep staging, pointing toward health-monitoring applications.
The paper releases its full code pipeline as open-source software, allowing independent labs to test the method on new data.
The per-class results imply that real deployments would need many more examples of rare emotions before a seven-way emotion classifier becomes trustworthy.

Reading between the lines

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

Editorial inference: The reported dataset cannot rule out that the network is detecting correlated physical activity rather than felt emotion; a blinded, physically separated protocol would separate the two.
Editorial inference: Taking the per-class results at face value, the practical signal is a coarse calm-versus-aroused or pleasant-versus-unpleasant axis, not fine-grained seven-way emotion recognition.
Editorial inference: The evolutionary story predicts a testable gradient: strongly defended plant species should show weak or absent responses and should fail the same classifier.
Editorial inference: The 'electromagnetic' framing is one of several possible channels; voice, breathing, or temperature cues could produce the same data, so targeted occlusion experiments (soundproofing, filtered airflow, blocked sightlines) could identify the actual mechanism.

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.

Desk Editor's Note

read the letter

You should know before reading: the abstract says 97% accuracy on seven emotions, but Table 1 shows three of the seven classes are never predicted and macro F1 is 0.56. The number is just accuracy on the majority classes (Neutral and Sad are 81% of the test set). That alone sinks the central claim as stated.

What is genuinely new here is the evolutionary wrapper, the "early warning hypothesis" for anti-herbivory detection. The empirical core, however, is not new: the ResNet50 emotion classification from plant spectrograms is essentially a repeat of the authors' own 2024 Sensors paper (ref [11]). This manuscript adds context and a narrative, not new measurements.

Credit where it is due: the methods section is concrete, the code is on GitHub, and the limitation section openly says the mechanism is unknown. The species-response hierarchy (basil responding more than defended species) is at least a coherent, testable pattern. Those are real features.

The soft spots are load-bearing. First, no blinding or physical separation is reported for the induction sessions here; the participant and plant are in the same room, so the model may be detecting body movement, speech, or electrode drift rather than emotion. The Faraday-cage artifact control is only cited from prior work, not shown. Second, the shuffled-label control hits 30%, which is below the 51% majority-class baseline; that means the control model isn't the same decision rule, so "97% versus 30%" is not a clean significance contrast. Third, three emotion classes have zero samples correctly classified, so the "seven-way decoding" language is simply inaccurate.

Who is this for? Anyone tracking the plant-electrophysiology fringe, or interested in how ML can produce optimistic numbers from small, unbalanced datasets. It is a useful teaching case, but not a reliable scientific claim. A serious editor could send it to review because the topic is high-impact-if-true and the code is available for auditors, but my own verdict is that it fails on its own reported numbers. I would not cite it for any substantive claim.

Referee Report

4 major / 6 minor

Summary. The paper reports a five-year research program claiming that plants generate bioelectric signals encoding human proximity, identity, movement, voice, stress, and emotion. The central new result is a ResNet50 classifier applied to mel-spectrograms of basil voltage recordings that allegedly achieves 97% accuracy in distinguishing seven human emotional states, with a shuffled-label control reported at 30% accuracy. The authors interpret these results as evidence for an evolved anti-herbivory early warning system and propose applications in agriculture, healthcare, and human-plant interaction. The methods describe an ESP32/INA128 sensor, 20-second spectrogram windows, and transfer learning with ResNet50, with code and data links provided.

Significance. If the 97% emotion-classification result were valid and causally attributable to the participant's internal emotional state, it would be a remarkable finding with implications for plant sensory biology and human-plant interaction. The paper deserves credit for releasing open-source code, for documenting a multi-year research program, and for attempting machine-learning-based analysis of plant bioelectric signals. However, as presented, the central claim is not supported: the confusion matrix shows that three of seven emotion classes are never predicted, macro F1 is 0.56, the shuffled-label control is not a valid chance baseline, and the experimental design does not exclude the most salient confound of concurrent human activity. The significance of the claimed phenomenon is therefore not established by this manuscript.

major comments (4)

[§4.1, Table 1] The headline 97% accuracy does not represent seven-way emotion classification. In Table 1, Disgusted, Fearful, and Surprised all have precision, recall, and F1 of 0.00, and the macro F1 is 0.56. The two majority classes, Neutral (462) and Sad (261), account for 723 of 898 test samples, so a trivial classifier that always predicts these two classes would achieve 80.5% accuracy. The 97% figure is therefore dominated by majority classes, and the abstract's claim of 'classifying human emotional states' is misleading without reporting per-class metrics and a majority-class baseline.
[§3.1, §4.4] The experimental design does not control for the most plausible confound: the participant and the plant are in the same room during emotion induction, and the electrode chain is sensitive to low-frequency electrical activity in the 0.1-50 Hz band. No Faraday cage, physical separation, audio masking, or blinded electrode setup is reported for this study; the artifact controls cited in §4.4 are attributed to 'our previous artifact elimination work' and no data are shown here. Under these conditions, the classifier may be detecting speech, facial movement, breathing, body shifts, or electrode drift correlated with the emotion-induction procedure rather than with the participant's internal emotional state. This is a load-bearing internal-validity gap, not a presentation issue.
[§4.1, Table 2] The shuffled-label control is not a valid null baseline as interpreted. With the reported class frequencies, random guessing according to empirical priors gives an expected accuracy of about 36.5%, and always predicting the majority class gives 51.4%. The shuffled model's 30% accuracy is below both, indicating that the shuffled training run did not implement the same decision rule as the valid model. The contrast '97% versus 30%' therefore does not establish that plant voltage spectrograms contain emotion information; a proper permutation control using the identical pipeline and decision rule is required.
[§4.4, §6.1] The manuscript asserts 'geographic replication confirmed in laboratories across Europe and North America,' 'seasonal consistency,' and 'rigorous artifact elimination' but provides no data, protocols, or effect sizes for these claims in this paper. Since the Discussion and Conclusions rely on these validations, the unsupported claims cannot be accepted as evidence. In addition, the Limitations section (6.1) acknowledges that the mechanism is unknown but does not acknowledge the confound-control or class-imbalance problems that directly bear on the main claim, making the limitations statement incomplete.

minor comments (6)

[§3.1] The Methods do not report the number of participants, number of sessions, demographic information, or emotion-induction procedure details; these are needed for reproducibility and for assessing the generalizability of the classifier.
[§3.3] The 80/20 train/test split is described as stratified, but it is not stated whether the split was performed at the level of individual windows, sessions, or participants. Because segmentation uses 20-second windows with 10-second overlap, a window-level split could leak information from the same session into both training and test sets; this should be clarified.
[§4.3] The text states that the primary discriminative signal is in the 0.5-10 Hz range, while the preprocessing section reports a 0.1-50 Hz bandpass filter; the paper should clarify whether classification used the full filtered band or a narrower sub-band.
[Figure 3] The caption states that the system 'achieved 97% confidence in detecting happiness,' which is a face-api.js confidence score for facial expression, not the plant classifier's confidence; the wording should distinguish these two very different quantities.
[§5.2, Reference [14]] The Codariocalyx motorius acoustic-sensing study is cited as a supporting evidence source, but it is a Master's thesis; the text should describe its methods and controls rather than relying solely on the citation.
[Throughout] There are several typographical issues, including 'Phase2from2020-2022' in §1.1, 'T raining Protocol' and 'V alidation Strategy' in §3.3, and inconsistent capitalization of 'Codariocalyx motorius' versus 'Codariocalyx Motorius' in §5.2 and reference [14].

Circularity Check

1 steps flagged · score 4.0 of 10

The emotion-classification result is an empirical supervised-learning result with independent code, but its interpretation as genuine plant bioelectric signaling rests on a load-bearing self-citation for Faraday-cage artifact control.

self citation load bearing [Section 4.4 (Environmental and Methodological Validation)]
"Building on our previous artifact elimination work, artifact control results showed effects persisting in Faraday cage conditions and responses maintained on isolation platforms. Temperature and humidity variations did not explain signal patterns, and randomized timing eliminated circadian artifacts."

The paper's central conclusion that the 97% emotion-classification accuracy reflects genuine plant bioelectric early-warning signals, rather than the classifier detecting correlated human activity during emotion induction, depends on ruling out such artifacts. Section 4.4 rules them out by asserting 'artifact control results showed effects persisting in Faraday cage conditions,' but the only justification is 'our previous artifact elimination work'—a self-citation to the authors' prior studies. No Faraday-cage data, protocol, or statistics for the present emotion sessions are shown.

full rationale

The paper's primary new result—97% accuracy in a ResNet50 classifier on held-out plant-voltage spectrograms—is an empirical machine-learning result with open-source code and data availability, and no equation in the paper defines that accuracy in terms of the inputs in a way that would make it circular. The 97% figure is also a reported test-set performance, not a fitted parameter renamed as a prediction. However, the paper's interpretation that this accuracy evidences 'genuine biological phenomena' rather than recording artifacts depends on §4.4, which cites only the authors' previous artifact-elimination work for Faraday-cage and isolation-platform controls; that control evidence is not presented in this paper. This is a load-bearing self-citation, though the central classification result still has independent content. I do not treat the Table 1/Table 2 discrepancies (three emotions never predicted; shuffled-label accuracy below the majority-class baseline) as circularity; those are statistical-validity and reporting concerns, not instances where the derivation reduces to its own inputs by construction.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The paper's central claim is the classification accuracy, which is wholly determined by the trained model parameters. The spectrogram settings are hand-chosen. No independent physical constants are involved, and the early warning hypothesis introduces no new entities, only a reinterpretation of existing phenomena.

free parameters (2)

ResNet50 model weights = trained on 80% of the plant voltage data
The claimed 97% accuracy is entirely determined by these fitted weights on the same data distribution; no independent test of generalization is provided.
Spectrogram hyperparameters (n_fft=256, hop length=64, n_mels=64) = hand-chosen
These choices affect the feature representation and are not derived from biological constraints.

assumptions (3)

domain assumption The plant electrode setup measures genuine bioelectric signals rather than electrochemical artifacts from the electrodes.
The paper states electrodes were attached to leaves and soil, but it does not provide impedance checks or control for electrode polarization during the emotion sessions. Section 4.3 mentions electrode polarization as a long-term effect, not a per-trial confound.
domain assumption Emotional induction through videos reliably produces the intended discrete emotions in all participants.
Section 3.1 mentions 'standardized emotional induction procedures' and face-api.js validation, but gives no participant self-report or validation accuracy.
ad hoc to paper The shuffled-label control is a valid baseline for testing information content.
A model trained on shuffled labels should achieve at least the majority-class accuracy (~51%), but the reported 30% is far below that, so the control does not behave as expected and its use as a null model is questionable.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Plant Bioelectric Early Warning Systems: A Five-Year Investigation into Human-Plant Electromagnetic Communication." pith.science (2026). https://pith.science/paper/UG2HOG22

@misc{pith2026250604132,
  author       = {Pith},
  title        = {Pith review of: Plant Bioelectric Early Warning Systems: A Five-Year Investigation into Human-Plant Electromagnetic Communication},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UG2HOG22}},
  note         = {Machine review of arXiv:2506.04132}
}

read the original abstract

We present a comprehensive investigation into plant bioelectric responses to human presence and emotional states, building on five years of systematic research. Using custom-built plant sensors and machine learning classification, we demonstrate that plants generate distinct bioelectric signals correlating with human proximity, emotional states, and physiological conditions. A deep learning model based on ResNet50 architecture achieved 97% accuracy in classifying human emotional states through plant voltage spectrograms, while control models with shuffled labels achieved only 30% accuracy. This study synthesizes findings from multiple experiments spanning 2020-2025, including individual recognition (66% accuracy), eurythmic gesture detection, stress prediction, and responses to human voice and movement. We propose that these phenomena represent evolved anti-herbivory early warning systems, where plants detect approaching animals through bioelectric field changes before physical contact. Our results challenge conventional understanding of plant sensory capabilities and suggest practical applications in agriculture, healthcare, and human-plant interaction research.

Figures

Figures reproduced from arXiv: 2506.04132 by the authors.

**Figure 2.** Circuit schematic of the ESP32-based plant bioelectric measurement system. The INA128 in [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Real-time emotion detection system showing facial expression analysis with confidence scores. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

16 extracted references · 16 canonical work pages

[11]

A., Ciechanowski, L., Dupuis, A., Vazquez, I., & Gloor, P

Kruse, J. A., Ciechanowski, L., Dupuis, A., Vazquez, I., & Gloor, P. A. (2024). Leveraging the sensitivity of plants with deep learning to recognize human emotions.Sensors, 24(6), 1917

work page 2024
[1]

K., & Gloor, P

Bhave, A., Renold, F. K., & Gloor, P. A. (2024). Using plants as biosensors to measure the emotions of jazz musicians. Handbook of Social Computing , 173-188

work page 2024
[2]

Burdon-Sanderson, J. (1873). Note on the electrical phenomena which accompany irritation of the leaf of Dionaea muscipula. Proceedings of the Royal Society of London , 21, 495-496. 9

work page
[3]

A., & Weinbeer, M

De La Cal, L., Gloor, P. A., & Weinbeer, M. (2023). Can plants sense humans? Using plants as biosensors to detect the presence of eurythmic gestures.Sensors, 23(15), 6971

work page 2023
[4]

Duerr, S., van Delden, J., Oezkaya, B., & Gloor, P. A. (2020). Eurythmic dancing with plants– measuring plant response to human body movement in an anthroposophic environment.arXiv preprint arXiv:2012.12978

work page Pith review arXiv 2020
[5]

Fromm, J., & Lautner, S. (2007). Electrical signals and their physiological significance in plants.Plant, Cell & Environment , 30(3), 249-257

work page 2007
[6]

I., & Gloor, P

Fuchs, P., von der Grün, R., Maslatón, C. I., & Gloor, P. A. (2024). Plants as biosensors: tomato plants’ reaction to human voices.Handbook of Social Computing , 294-309

work page 2024
[7]

Gagliano, M., Renton, M., Depczynski, M., & Mancuso, S. (2014). Experience teaches plants to learn faster and forget slower in environments where it matters.Oecologia, 175(1), 63-72

work page 2014

Show all 16 references

[8]

Gallé, A., Lautner, S., Flexas, J., & Fromm, J. (2015). Environmental stimuli and physiological re- sponses: the current state of the art in plant electrophysiology. Journal of Plant Research , 128(2), 193-209

2015
[9]

Geddes, L. A. (1972).Electrodes and the measurement of bioelectric events . Wiley-Interscience

1972
[10]

F., Weinbeer, M., & Gloor, P

Gil, A. F., Weinbeer, M., & Gloor, P. A. (2024). Can plants perceive human gestures? Using AI to track eurythmic human–plant interaction.Biomimetics, 9(5), 290

2024
[12]

R., Bikson, M., & Jefferys, J

Merrill, D. R., Bikson, M., & Jefferys, J. G. (2005). Electrical stimulation of excitable tissue: design of efficacious and safe protocols.Journal of Neuroscience Methods , 141(2), 171-198

2005
[13]

Oezkaya, B., & Gloor, P. A. (2020). Recognizing individuals and their emotions using plants as bio- sensors through electro-static discharge.arXiv preprint arXiv:2005.04591

2020 arXiv
[14]

Peter, P. K. (2021). Do Plants sense music? An evaluation of the sensorial abilities of theCodariocalyx Motorius. Masters thesis, Universität zu Köln. Available at: https://kups.ub.uni-koeln.de/53756/

2021
[15]

O., & Gloor, P

Smolin, D., Çelik, D. O., & Gloor, P. A. (2025). Predicting human stress and exam performance through variations of the electrical potentials of a basil plant.COINs2025 Proceedings

2025
[16]

Tabuenca, B., Uche-Soria, M., Greller, W., Hernández-Leo, D., Balcells-Falgueras, P., Gloor, P., & Gar- bajosa, J. (2024). Greening smart learning environments with Artificial Intelligence of Things.Internet of Things, 25, 101051. 10

2024

Pith tools

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

[1] [11]

A., Ciechanowski, L., Dupuis, A., Vazquez, I., & Gloor, P

Kruse, J. A., Ciechanowski, L., Dupuis, A., Vazquez, I., & Gloor, P. A. (2024). Leveraging the sensitivity of plants with deep learning to recognize human emotions.Sensors, 24(6), 1917

work page 2024

[2] [1]

K., & Gloor, P

Bhave, A., Renold, F. K., & Gloor, P. A. (2024). Using plants as biosensors to measure the emotions of jazz musicians. Handbook of Social Computing , 173-188

work page 2024

[3] [2]

Burdon-Sanderson, J. (1873). Note on the electrical phenomena which accompany irritation of the leaf of Dionaea muscipula. Proceedings of the Royal Society of London , 21, 495-496. 9

work page

[4] [3]

A., & Weinbeer, M

De La Cal, L., Gloor, P. A., & Weinbeer, M. (2023). Can plants sense humans? Using plants as biosensors to detect the presence of eurythmic gestures.Sensors, 23(15), 6971

work page 2023

[5] [4]

Duerr, S., van Delden, J., Oezkaya, B., & Gloor, P. A. (2020). Eurythmic dancing with plants– measuring plant response to human body movement in an anthroposophic environment.arXiv preprint arXiv:2012.12978

work page Pith review arXiv 2020

[6] [5]

Fromm, J., & Lautner, S. (2007). Electrical signals and their physiological significance in plants.Plant, Cell & Environment , 30(3), 249-257

work page 2007

[7] [6]

I., & Gloor, P

Fuchs, P., von der Grün, R., Maslatón, C. I., & Gloor, P. A. (2024). Plants as biosensors: tomato plants’ reaction to human voices.Handbook of Social Computing , 294-309

work page 2024

[8] [7]

Gagliano, M., Renton, M., Depczynski, M., & Mancuso, S. (2014). Experience teaches plants to learn faster and forget slower in environments where it matters.Oecologia, 175(1), 63-72

work page 2014