---
title: 'TinyMyo: Compact EMG Foundation Model'
url: https://www.emergentmind.com/topics/tinymyo
type: topic
---

# TinyMyo: Compact EMG Foundation Model

Searching arXiv for the cited TinyMyo-related papers and adjacent work to ground the article in the provided literature.
TinyMyo most commonly denotes a tiny foundation model for surface electromyography (EMG): a lightweight Transformer encoder pretrained in a self-supervised manner on heterogeneous public EMG datasets and then adapted with minimal task-specific heads to multiple downstream problems, including hand gesture classification, hand kinematic regression, speech production, and speech recognition [2512.15729]. Its stated purpose is to address poor generalization across subjects, recording systems, and acquisition protocols while remaining small enough for edge deployment, including deployment on the ultra-low-power GAP9 microcontroller. In the supplied literature, the same name also appears in a distinct biophysical context as a Python-based automated tracking workflow for myosin II filaments, so the term is not globally unambiguous [2001.10035].

## 1. Scope, problem setting, and nomenclature

TinyMyo was introduced against a familiar EMG systems problem: surface EMG is non-invasive and useful in biomechanics, rehabilitation, prosthetic control, and human-machine interaction, but the signals vary because of noise, motion artifacts, cross-talk, inter-subject and intra-subject variability, and differences in electrode layouts, sampling rates, amplification, filtering, devices, and acquisition protocols [2512.15729]. The work positions these factors as the main obstacle to transferable EMG modeling.

Within that framing, TinyMyo is defined as a **single pretrained EMG backbone** that can be reused across multiple downstream tasks. The paper explicitly contrasts this with conventional supervised models that are usually task-specific, and with prior EMG foundation models that tend to be large, limited to one downstream task, or not deployable on embedded platforms. The backbone is therefore intended to serve as a common representation layer rather than as a single-purpose decoder.

A recurrent misconception is to treat TinyMyo as simply a small gesture classifier. The supplied literature does not support that reduction. The model is evaluated on classification, regression, speech production, and speech recognition, and the authors frame it as a **flexible resource** rather than a one-task architecture. A second source of confusion is terminological: in a separate paper summary, “TinyMyo” refers to a tracking routine for myosin II filaments rather than to EMG processing, which is a genuinely different use of the same label [2001.10035].

## 2. Transformer architecture and self-supervised pretraining

TinyMyo is a **Transformer encoder foundation model operating directly on time-domain EMG**. The backbone has **8 pre-LayerNorm Transformer blocks**, **embedding dimension 192**, **3 attention heads**, and about **3.6M parameters**. The pretraining decoder is deliberately tiny, at only **3.9k parameters**, so that most representational burden remains in the encoder [2512.15729].

The input representation uses a **channel-independent patching strategy**. For an input waveform
$$
\mathbf{X}\in\mathbb{R}^{T\times C},
$$
TinyMyo splits each channel separately into temporal patches,
$$
\mathbf{P}\in\mathbb{R}^{C\times N_p\times L}, \qquad N_p=\left\lfloor\frac{T-L}{S}\right\rfloor+1.
$$
During pretraining, the paper uses \(T=1000\) samples, \(L=S=20\), and \(C=16\), yielding a total sequence length of \(N=800\). Each patch is projected by a shared linear map, and positional structure is handled with **RoPE** rather than learned positional embeddings. The authors explicitly justify the channel-independent tokenization by noting that each channel corresponds to a distinct anatomical electrode site, so inter-channel mixing is deferred to attention rather than imposed at tokenization time.

Pretraining uses **masked reconstruction** with a **50% mask ratio**. The losses are
$$
\mathcal{L}_{\mathrm{masked} = \frac{1}{|\mathcal{MASK}|} \sum_{(c,i)\in\mathcal{MASK} \mathrm{SmoothL1}\big(\mathbf{P}_{c,i},\hat{\mathbf{P}_{c,i}\big),
$$
$$
\mathcal{L}_{\mathrm{visible} = \frac{1}{|\mathcal{M}|} \sum_{(c,i)\in\mathcal{M} \mathrm{SmoothL1}\big(\mathbf{P}_{c,i},\hat{\mathbf{P}_{c,i}\big),
$$
and
$$
\mathcal{L}_{\mathrm{total} = \mathcal{L}_{\mathrm{masked} + 0.1\cdot \mathcal{L}_{\mathrm{visible}.
$$
The downweighting of the visible-patch term is used to reduce trivial copying.

The pretraining corpus consists of **NinaPro DB6**, **NinaPro DB7**, and **EMG2Pose**. The summaries specify **10 subjects** and **14 channels** for DB6, **22 subjects** and **12 channels** for DB7, and **193 participants** and **16 channels** for EMG2Pose. Preprocessing uses a **Butterworth band-pass filter \(20{-}450\) Hz**, a **50 Hz notch filter**, and **channel-wise min-max normalization to \([-1,1]\)**. Windows have **1000 samples = 500 ms** with **50% overlap**; channels fewer than 16 are zero-padded; **no data augmentation** is used; optimization uses **AdamW**, **batch size 512**, **peak LR \(10^{-4}\)**, and **50 epochs with 10 warmup epochs**. Training was performed on **CSCS Alps with NVIDIA GH200 GPUs in DDP mode** [2512.15729].

## 3. Downstream interface, benchmarks, and empirical results

After pretraining, the decoder is discarded and the encoder is reused with **minimal task-specific heads**. Channel-wise embeddings are fused by **concatenation rather than averaging**,
$$
[\mathbf{z}_{1,p}\Vert\cdots\Vert\mathbf{z}_{C,p}],
$$
and then temporally average-pooled before the task head. The paper reports that concatenation preserved electrode-specific information and empirically outperformed mean fusion [2512.15729].

The downstream evaluation spans four task families and several datasets acquired with different sensing locations and hardware platforms.

| Task family | Dataset(s) | Reported result |
|---|---|---|
| Hand gesture classification | NinaPro DB5, EPN-612, UCI-EMG | \(89.41 \pm 0.16\%\), \(96.74 \pm 0.09\%\), \(97.56 \pm 0.32\%\) |
| Hand kinematic regression | NinaPro DB8 | MAE \(8.77 \pm 0.12^\circ\), RMSE \(13.35 \pm 0.15^\circ\), \(R^2 = 0.62 \pm 0.01\) |
| Speech production | Gaddy Silent Speech Dataset | WER \(33.54 \pm 1.12\%\) |
| Speech recognition | Gaddy Silent Speech Dataset | WER \(33.95 \pm 0.97\%\) |

For **hand gesture classification**, the headline results are reported as state of the art compared to previous FM-based work on **NinaPro DB5**, **UCI-EMG**, and **EPN-612**. On **NinaPro DB5**, TinyMyo reports **FS, 200 ms: \(81.94 \pm 1.22\%\)** and **FT, 200 ms: \(89.41 \pm 0.16\%\)**, with the latter marked as the best result in the table. On **EPN-612**, it reports **FS, 1000 ms: \(95.18 \pm 0.10\%\)** and **FT, 1000 ms: \(96.74 \pm 0.09\%\)**. On **UCI-EMG**, the paper notes that the task is simple enough that pretraining offers little extra benefit, with **FS, 1000 ms: \(97.56 \pm 0.32\%\)** and **FT, 1000 ms: \(97.23 \pm 0.70\%\)**.

The paper also introduces the **Generic Neuromotor Interface** as a new downstream benchmark for EMG foundation models. There, **TinyMyo FS** achieves **CLER \(0.144\pm0.004\)** and **TinyMyo FT** achieves **CLER \(0.153\pm0.006\)**, while the original LSTM baselines are **0.1596** for windowed inference and **0.1819** for full sequence. The authors attribute the weaker FT result to a mismatch between TinyMyo’s bidirectional pretraining and the benchmark’s causal/windowed inference constraint.

For **hand kinematic regression** on **NinaPro DB8**, the best reported result is **TinyMyo FT, 1000 ms**, with **MAE \(8.77 \pm 0.12^\circ\)**, **RMSE \(13.35 \pm 0.15^\circ\)**, and **\(R^2 = 0.62 \pm 0.01\)**. The paper explicitly notes that comparisons with subject-specific prior work such as TEMPONet TCN or event-based linear regression are not directly apples-to-apples, because TinyMyo is evaluated in an across-subject setting.

For **speech production**, the model is used in a three-stage pipeline: EMG-to-MFCC transduction, **HiFi-GAN** vocoding, and ASR-based transcription for evaluation. The transduction model uses **3 residual convolutional blocks**, downsamples from **1600 to 200 samples**, and predicts **26 MFCCs**. On the **Gaddy Silent Speech Dataset**, TinyMyo reports **FS: \(35.23 \pm 0.22\%\)** and **FT: \(33.54 \pm 1.12\%\)** WER. The transduction component is about **4.5M parameters**, compared with **54M** in the Gaddy baseline, described as roughly **91.7% reduction** in model size.

For **speech recognition**, the same dataset is used with a CTC-based head, a **4-gram language model on LibriSpeech**, and beam search decoding. The reported result is **TinyMyo FT: \(33.95 \pm 0.97\%\)** WER. The paper presents this as a smaller, more deployable EMG-only alternative to larger or multimodal systems such as MONA and MONA LISA [2512.15729].

## 4. Edge deployment and systems profile

A central claim of TinyMyo is deployability at the edge. The paper reports, to the best of its knowledge, the **first deployment of an EMG foundation model on an ultra-low-power microcontroller**, specifically **GreenWaves GAP9** [2512.15729].

The deployment target has **external L3 HyperRAM**, **1.5 MB on-chip L2**, and **128 kB L1**. Because attention has \(O(N^2)\) scaling and the deployed gesture model uses sequence length **800**, the implementation relies on a **hierarchical streaming toolchain**: weights and activations reside in L3, slabs are streamed into L2 on demand, tiles are moved from L2 to L1 for computation, and **double buffering** is used to hide transfer latency. The summaries also note **offline liveness analysis**, **static memory arena allocation**, and integer-only implementations of **softmax**, **LayerNorm**, and **GELU**.

Quantization is **INT8** for **weights**, **activations**, **MHSA**, **GELU**, and **LayerNorm**. Only **LayerNorm scale/shift** remain **FP32**, and the final classification layer accumulates to **FP32 logits** for comparison. The per-block computation profile is dominated by attention and MLP projections: **Q/K/V projections: 88M MACs**, **QK scores: 123M MACs**, **AV context: 123M MACs**, **output projection: 29M MACs**, **FC1: 118M MACs**, and **FC2: 118M MACs**.

The reported deployment result on GAP9 at **370 MHz** using **all 9 cores** is **~12.2 s** inference time, **0.44 J** energy, and **36.45 mW** average power envelope. These numbers establish deployability, but the paper also explicitly notes a limitation: **inference is still too slow for seamless real-time use**. That caveat is important. The work demonstrates feasibility under severe power constraints, not completion of the latency problem.

## 5. Relation to compact wearable EMG literature

TinyMyo belongs to a longer line of efforts to make EMG inference compact, accurate, and deployable. Earlier work on **“Compact Deep Neural Networks for Computationally Efficient Gesture Classification From Electromyography Signals”** proposed a compact SqueezeNet-inspired CNN with a customized **Temporal Fire Module** for **15-class hand gesture recognition** from Myo and Delsys recordings. On the **Myo Armband**, that model reported **84.2 ± 6%** accuracy versus **70.5 ± 7%** for an SVM, with only **5,889 parameters** and **7.89 ms** inference on **Jetson TX2** [1806.08641]. In historical terms, that paper established that deep EMG models could be much smaller than prior alternatives without relinquishing embedded relevance.

A parallel line emphasized direct wearable control rather than representation learning. **“Teleoperated Robotic Arm Movement Using EMG Signal With Wearable MYO Armband”** used the wireless **Myo gesture armband** on the forearm, with **8 sEMG sensors**, **200 Hz** sampling, time-domain features **MAV, WL, RMS, AR, ZC, and SSC**, and a command pipeline for a **5-DoF robotic arm**. The reported offline accuracies were **SVM: 96.57%**, **LDA: 96.01%**, and **KNN: 92.67%**, with optimized online-mode performance of **SVM: 95.27%**, **LDA: 94.53%**, and **KNN: 89.43%** [1810.09929]. This work showed that Myo-class hardware could support stable real-time teleoperation with classical pattern-recognition methods.

More recent work shifts attention from compactness alone to **session drift** and long-term usability. **“Lightweight Test-Time Adaptation for EMG-Based Gesture Recognition”** uses a compact **Temporal Convolutional Network** with about **47k parameters** and reports that a baseline model achieves about **85.12% intra-session accuracy** but only **56.61% inter-session accuracy** on **NinaPro DB6**. Three lightweight adaptation strategies then improve inter-session performance: **causal adaptive BatchNorm** to about **68.93%**, **GMM + alignment** to about **69.78%**, and **meta-learning** to about **80.30%** [2601.04181]. A plausible implication is that TinyMyo’s pretrained backbone and lightweight test-time adaptation are complementary rather than competing ideas: one addresses transferable representation learning, the other addresses deployment-time non-stationarity.

Against this background, TinyMyo can be situated as a shift from **task-specific or session-specific decoders** toward a **shared, pretrained, edge-aware EMG backbone**. Its novelty is not merely smaller size, although **3.6M parameters** remains modest by foundation-model standards; it is the combination of multi-task reuse, cross-device transfer, and demonstrated MCU deployment [2512.15729].

## 6. Distinct biophysical use of the name

In a separate literature, TinyMyo refers not to EMG modeling but to a **Python-based automated single-particle tracking workflow** for **myosin II filaments** in **actin networks tethered to supported lipid bilayers** [2001.10035]. This workflow was demonstrated on **interferometric scattering microscopy (iSCAT)** videos and was designed to track both **position** and **orientation**, and to classify motion into **diffusive** and **processive** segments.

The pipeline has three stages: detection of filament-like objects, linking detections into tracks, and classifying track segments by displacement correlations and angular motion. Detection uses **SEP** (derived from **SExtractor**) and accepts a candidate if **area > 40 pixels** and each pixel intensity is at least **1.5 \(\sigma\)** above local background. Two area cutoffs are then applied: a minimum reliable area of
$$
A_{\min} = 0.151 \,\mu\text{m}^2
$$
and a maximum single-filament cutoff of
$$
A_{\max} = 4A_{\text{av} = 3.251 \,\mu\text{m}^2.
$$
Tracks are linked with a centroid-displacement constraint
$$
|\Delta \mathbf{x}| \le 6 \text{ pxl},
$$
an area-consistency rule, and a minimum duration of **at least two frames**.

Its motion-state classifier uses a **2 s** dwell-time criterion and displacement-correlation thresholds derived from the measured average diffusivity
$$
D_{\text{avg} = (0.00048 \pm 0.00004)\,\mu\text{m}^2\,\text{s}^{-1}.
$$
Processive points are accepted only in regions of at least **five consecutive processive points** over about **1 s**. Orientation handling extends the angle domain beyond SEP’s \([-\pi/2,\pi/2]\) output so that angular diffusion can be treated continuously. The paper reports high-throughput analysis with **15,647 filaments** in one condition and **14,016 filaments** in another, and states that fewer than **4%** of tracks ended within 6 pixels of the image boundary.

This biophysical TinyMyo is therefore unrelated to the EMG foundation model except in the broad methodological sense that both are compact, automated, and intended for high-throughput or edge-constrained analysis. For disambiguation, the context is decisive: in EMG and wearable computing, TinyMyo denotes the Transformer-based foundation model; in acto-myosin microscopy, it denotes the myosin II filament tracking workflow.

Source: https://www.emergentmind.com/topics/tinymyo