--- title: "ArEEG_Chars: Dataset for Envisioned Speech Recognition using EEG for Arabic — Darwish 2024" description: "The authors built ArEEG_Chars, which they describe as the first such Arabic resource. 92.6% F1. 13 references." authors: ["Hazem Darwish", "Abdalrahman Al Malah", "Khloud Al Jallad", "Nada Ghneim"] year: 2024 venue: "arXiv:2402.15733" source: "https://arxiv.org/abs/2402.15733" doi: "10.48550/arxiv.2402.15733" arxiv: "2402.15733" licence: "CC BY-NC-ND 4.0 (arXiv)" licence_url: "http://creativecommons.org/licenses/by-nc-nd/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/darwish-2024-areeg-chars-dataset-envisioned-speech/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/darwish-2024-areeg-chars-dataset-envisioned-speech/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2024](https://md.research.elamur.ai/free/en/papers/2024/page/5/) / ArEEG\_Chars: Dataset for Envisioned Speech Recognition using EEG for Arabic Characters # ArEEG\_Chars: Dataset for Envisioned Speech Recognition using EEG for Arabic Characters **Authors:** [Hazem Darwish](https://md.research.elamur.ai/free/en/authors/hazem-darwish/), [Abdalrahman Al Malah](https://md.research.elamur.ai/free/en/authors/abdalrahman-al-malah/), [Khloud Al Jallad](https://md.research.elamur.ai/free/en/authors/khloud-al-jallad/) and [Nada Ghneim](https://md.research.elamur.ai/free/en/authors/nada-ghneim/) · **Year:** 2024 · **Venue:** arXiv:2402.15733 · **Licence:** [CC BY-NC-ND 4.0 (arXiv)](http://creativecommons.org/licenses/by-nc-nd/4.0/) **Links:** [Publisher's page](https://arxiv.org/abs/2402.15733) · [PDF](https://arxiv.org/pdf/2402.15733) · [DOI 10.48550/arxiv.2402.15733](https://doi.org/10.48550/arxiv.2402.15733) · [arXiv 2402.15733](https://arxiv.org/abs/2402.15733) · [OpenAlex](https://openalex.org/W4392223715) **Citations:** 2 citations (OpenAlex, 2026-10-14) ## Summary The paper targets a gap in brain-computer interface research: no public EEG data existed for Arabic characters, while English letters, words, and directions already had several benchmarks. The authors built ArEEG_Chars, which they describe as the first such Arabic resource. Thirty native speakers wore a 14-channel Emotiv EPOC X headset. Each viewed every one of the 31 Arabic letters on a slide for 10 seconds, then imagined it with eyes closed for 10 seconds, with a 20-second rest between letters. Slides were shuffled per session, and participants were asked at several points whether their focus had slipped. This gave 930 recordings, which were cut into 250 ms windows to yield 39,857 segments. The dataset is released on Mendeley Data. The paper also surveys earlier imagined-speech datasets and the preprocessing, feature, and classifier pipelines used on them. No classification results are reported; baseline deep learning models are left as future work. The resource matters because it opens envisioned-speech BCI research to Arabic speakers, including people with communication disabilities. ## Key points - **Dataset:** ArEEG_Chars covers 31 Arabic characters from 30 participants (21 male, 9 female). Ages run roughly from the mid-teens to the early 60s, with a mean around 26. It is published on Mendeley Data (doi 10.17632/f2bgr7tdtt.1). - **Hardware:** Emotiv EPOC X, with 14 electrodes placed according to the 10-20 system (AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8, AF4) and CMS/DRL references. Signals were sampled at 2048 Hz, downsampled to 128 Hz, and sent over Bluetooth. - **Protocol:** - 10 seconds viewing each letter, then 10 seconds imagining it with eyes closed. - 20-second pause between letters. - Letter order randomized. - Participants abstained from caffeine, alcohol, and smoking beforehand. - Each session took about an hour per person. - **Volume:** 930 raw recordings were segmented into 39,857 windows of 250 ms. - **Related work:** The surveyed systems include Kumar et al. (Random Forest, 82.20% on digits, letters, and images), an LSTM on wavelet-scattering features for direction commands (about 92.6% F1), and SVM with CSP on two vowels (71%). ## Interesting ideas - Shuffling the letter slides for each session stops participants from anticipating the next stimulus, so order effects cannot masquerade as signal. - Asking participants at the start, middle, and end whether they lost focus is a cheap self-report check on label quality during collection. - A long rest between trials clears the previous mental image, so one trial does not contaminate the next. - Cutting each long recording into short fixed windows multiplies the number of training samples. The catch is that windows from one recording are strongly correlated, so splits should be made by recording or participant. - Logging per-contributor metadata (age, gender, education) makes stratified error analysis possible later. ## For Mova How these ideas could serve Mova, our small open language model — its files, stages and tests: - **I1** (learn): shuffle item order when querying LLM teacher tiers or running evaluations, so labels and scores do not depend on the sequence in which examples are presented. - **I2** (learn): have teacher tiers and witness parsers emit an explicit "unsure" flag per item, and keep flagged silver trees out of training or down-weight them. - **I3** (answer): rebuild the world of the text from scratch for each test question, and add a negative control that checks answers do not change when the preceding item is swapped. - **I4** (learn): when sentence-level silver data is derived from documents, split train and test by source document, so evaluation is not inflated by near-duplicate context. - **I5** (learn): store provenance for every silver annotation (which witness parsers agreed, genre, source), so failures can be traced to specific data slices. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* Fields: [Linguistics & languages](https://md.research.elamur.ai/free/en/shelves/linguistics/) Topics: [🎙️ Speech & spoken language](https://md.research.elamur.ai/free/en/topics/speech/) · [🐪 Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) ## Institutions - [Arab International University](https://md.research.elamur.ai/free/en/universities/arab-international-university/) — University, Ghabāghib, Syria · 8 works · 12 citations ## Cite Cite as: Hazem Darwish, Abdalrahman Al Malah, Khloud Al Jallad and Nada Ghneim. 2024. ArEEG\_Chars: Dataset for Envisioned Speech Recognition using EEG for Arabic Characters. arXiv:2402.15733. The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @misc{darwish2024areeg, title = {{ArEEG\_Chars: Dataset for Envisioned Speech Recognition using EEG for Arabic Characters}}, author = {Hazem Darwish, Abdalrahman {Al Malah}, Khloud {Al Jallad} and Nada Ghneim}, year = {2024}, doi = {10.48550/arxiv.2402.15733}, eprint = {2402.15733}, archivePrefix = {arXiv}, url = {https://doi.org/10.48550/arxiv.2402.15733} } ``` - [← previous AraTar: A Corpus to Support the Fine-grained Detection of Hate Speech Targets in the Arabic Language](https://md.research.elamur.ai/free/en/papers/alghamdi-2024-aratar-corpus-support-fine-grained/) - [next → ASOS at NADI 2024 shared task: Bridging Dialectness Estimation and MSA Machine Translation for Arabic Language Enhancement](https://md.research.elamur.ai/free/en/papers/nacar-2024-asos-nadi-2024-shared-task/) - [↑ back to papers of 2024](https://md.research.elamur.ai/free/en/papers/2024/page/5/) ## References 13 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Lorainne Tudor Car; Mario Tudor; Katarina Ivana Tudor, "Hans berger The history of electroencephalography," Acta medica Croatica: c̆ asopis Hravatske akademije medicinskih znanosti, p. 8, 2005. [find](https://scholar.google.com/scholar?q=Lorainne+Tudor+Car%3B+Mario+Tudor%3B+Katarina+Ivana+Tudor%2C+%22Hans+berger+The+history+of+electroencephalography%2C%22+Acta+medica+Croatica%3A+c%CC%86+asopis+Hravatske+akademije+medicinskih+znanosti%2C+p%2E+8%2C+2005%2E) 2. Pradeep Kumar, Rajkumar Saini, Partha Pratim Roy, Pawan Kumar Sahu, and Debi Prosad Dogra., "Envisioned speech recognition using EEG sensors.," Personal and Ubiquitous Computing., vol. 15, 2017. [find](https://scholar.google.com/scholar?q=Pradeep+Kumar%2C+Rajkumar+Saini%2C+Partha+Pratim+Roy%2C+Pawan+Kumar+Sahu%2C+and+Debi+Prosad+Dogra%2E%2C+%22Envisioned+speech+recognition+using+EEG+sensors%2E%2C%22+Personal+and+Ubiquitous+Computing%2E%2C+vol%2E+15%2C+2017%2E) 3. Nicolás Nieto; Victoria Peterson. Hugo Leonardo Rufner; Juan Esteban Kamienkowski; Ruben Spies, "Thinking out loud, an open-access EEG-based BCI dataset for inner speech recognition," Scientific data, p. 17, 2022. [find](https://scholar.google.com/scholar?q=Nicol%C3%A1s+Nieto%3B+Victoria+Peterson%2E+Hugo+Leonardo+Rufner%3B+Juan+Esteban+Kamienkowski%3B+Ruben+Spies%2C+%22Thinking+out+loud%2C+an+open%2Daccess+EEG%2Dbased+BCI+dataset+for+inner+speech+recognition%2C%22+Scientific+data%2C) 4. Mokhles M. Abdulghani; Wilbur L. Walters; Khalid H. Abed, "Imagined Speech Classification Using EEG and Deep Learning," bioengineering, p. 15, 2023. [find](https://scholar.google.com/scholar?q=Mokhles+M%2E+Abdulghani%3B+Wilbur+L%2E+Walters%3B+Khalid+H%2E+Abed%2C+%22Imagined+Speech+Classification+Using+EEG+and+Deep+Learning%2C%22+bioengineering%2C+p%2E+15%2C+2023%2E) 5. Siyi Deng; Ramesh Srinivasan., " EEG classification of imagined syllable rhythm using Hilbert spectrum methods.," Journal of Neural Engineering, p. 25, 2010. [find](https://scholar.google.com/scholar?q=Siyi+Deng%3B+Ramesh+Srinivasan%2E%2C+%22+EEG+classification+of+imagined+syllable+rhythm+using+Hilbert+spectrum+methods%2E%2C%22+Journal+of+Neural+Engineering%2C+p%2E+25%2C+2010%2E) 6. Aurélie de Borman; Benjamin Wittevrongel; Ine Dauwe; , Evelien Carrette; Alfred Meurs; Dirk Van Roost; Paul Boon; & Marc M. Van Hulle , "Imagined speech event detection from electrocorticography and its transfer between speech modes and subjects," communications biology, p. 12, 2024. [find](https://scholar.google.com/scholar?q=Aur%C3%A9lie+de+Borman%3B+Benjamin+Wittevrongel%3B+Ine+Dauwe%3B+%2C+Evelien+Carrette%3B+Alfred+Meurs%3B+Dirk+Van+Roost%3B+Paul+Boon%3B+%26+Marc+M%2E+Van+Hulle+%2C+%22Imagined+speech+event+detection+from+electrocorticography+and+i) 7. A. van Wieringen and J. Wouters, "LIST and LINT: sentences and numbers for quantifying speech understanding in severely impaired listeners for Flanders and the Netherlands," Int J. Audio, vol. 47, p. 348–355, 2008 . [find](https://scholar.google.com/scholar?q=A%2E+van+Wieringen+and+J%2E+Wouters%2C+%22LIST+and+LINT%3A+sentences+and+numbers+for+quantifying+speech+understanding+in+severely+impaired+listeners+for+Flanders+and+the+Netherlands%2C%22+Int+J%2E+Audio%2C+vol%2E+47%2C+p%2E+) 8. Charles S. DaSalla, Hiroyuki Kambara, Makoto Sato, Yasuharu Koike., "Single-trial classification of vowel speech imagery using common spatial patterns.," Neural Networks., p. 70, 2009. [find](https://scholar.google.com/scholar?q=Charles+S%2E+DaSalla%2C+Hiroyuki+Kambara%2C+Makoto+Sato%2C+Yasuharu+Koike%2E%2C+%22Single%2Dtrial+classification+of+vowel+speech+imagery+using+common+spatial+patterns%2E%2C%22+Neural+Networks%2E%2C+p%2E+70%2C+2009%2E) 9. Mariko Matsumoto; Junichi Hori, "Classification of silent speech using support vector machine and relevance vector machine," Applied Soft Computing, p. 20, 2014. [find](https://scholar.google.com/scholar?q=Mariko+Matsumoto%3B+Junichi+Hori%2C+%22Classification+of+silent+speech+using+support+vector+machine+and+relevance+vector+machine%2C%22+Applied+Soft+Computing%2C+p%2E+20%2C+2014%2E) 10. Alejandro A Torres-Garc´ıa; Carlos A. Reyes-Garc´ıa; Luis Villasenor-Pineda; Gregorio Garc´ıa-Aguilar, "Implementing a Fuzzy Inference System in a Multi-Objective EEG Channel Selection Model for Imagined Speech Classification," Expert Systems With Applications, p. 41, 2016. [find](https://scholar.google.com/scholar?q=Alejandro+A+Torres%2DGarc%C2%B4%C4%B1a%3B+Carlos+A%2E+Reyes%2DGarc%C2%B4%C4%B1a%3B+Luis+Villasenor%2DPineda%3B+Gregorio+Garc%C2%B4%C4%B1a%2DAguilar%2C+%22Implementing+a+Fuzzy+Inference+System+in+a+Multi%2DObjective+EEG+Channel+Selection+Model+for+Imagi) 11. Luis Alfredo Moctezuma; Alejandro A. Torres-Garc´ıa; Luis Villasenor-Pineda; Maya Carrillo-Ruiz., "Subjects Identification using EEG-recorded Imagined Speech," Expert Systems With Applications, p. 26, 2018. [find](https://scholar.google.com/scholar?q=Luis+Alfredo+Moctezuma%3B+Alejandro+A%2E+Torres%2DGarc%C2%B4%C4%B1a%3B+Luis+Villasenor%2DPineda%3B+Maya+Carrillo%2DRuiz%2E%2C+%22Subjects+Identification+using+EEG%2Drecorded+Imagined+Speech%2C%22+Expert+Systems+With+Applications%2C+p%2E+26%2C) 12. A. A. Torres-García, C. A. Reyes-Garcia and L. Villaseñor-Pineda, "Toward a silent speech interface based on unspoken speech," in Proceedings of BIOSIGNALS 2012 (BIOSTEC), Algarve, Portugal, 2018. [find](https://scholar.google.com/scholar?q=A%2E+A%2E+Torres%2DGarc%C3%ADa%2C+C%2E+A%2E+Reyes%2DGarcia+and+L%2E+Villase%C3%B1or%2DPineda%2C+%22Toward+a+silent+speech+interface+based+on+unspoken+speech%2C%22+in+Proceedings+of+BIOSIGNALS+2012+%28BIOSTEC%29%2C+Algarve%2C+Portugal%2C+2018%2E) 13. A. Gramfort, M. Luessi, E. Larson, D. A. Engemann, D. Strohmeier, C. Brodbeck, L. Parkkonen, M. S. Hämäläinen, , "MNE software for processing MEG and EEG data," NeuroImage, p. 86, 2014. [find](https://scholar.google.com/scholar?q=A%2E+Gramfort%2C+M%2E+Luessi%2C+E%2E+Larson%2C+D%2E+A%2E+Engemann%2C+D%2E+Strohmeier%2C+C%2E+Brodbeck%2C+L%2E+Parkkonen%2C+M%2E+S%2E+H%C3%A4m%C3%A4l%C3%A4inen%2C+%2C+%22MNE+software+for+processing+MEG+and+EEG+data%2C%22+NeuroImage%2C+p%2E+86%2C+2014%2E)