Meta AI Releases Brain2Qwerty v2: A Non-Invasive MEG Brain-to-Text Pipeline Decoding Typed Sentences at 61% Word Accuracy
English summary
Meta AI has released Brain2Qwerty v2, a non-invasive brain-to-text decoder that reconstructs typed sentences from magnetoencephalography (MEG) signals without surgery. The system achieves 61% average word accuracy (39% WER), a major jump from the 8% baseline of prior non-invasive methods; the best participant reached 78% accuracy with over half of sentences having at most one word error. The pipeline combines a convolutional encoder, transformer, and character-level language model, and it leverages fine-tuned large language models to incorporate semantic context. Accuracy improves log-linearly with the amount of MEG recording data. Meta has publicly released the full training code for both v1 and v2 under the CC BY-NC 4.0 license. This is a research release, tested on 9 healthy volunteers in a controlled lab setting; practical use is limited by the need for a magnetically shielded room and a still participant.
Chinese summary
Meta AI 发布了 Brain2Qwerty v2，这是一个非侵入式的脑信号到文本解码器，利用脑磁图（MEG）信号实时重建打字句子，无需手术植入。该系统平均词准确率达 61%（词错误率 39%），远高于此前非侵入式方法的 8% 基准，表现最佳的参与者达到了 78% 的准确率，一半以上的句子错误不超过一个词。该流程结合了卷积编码器、Transformer 和字符级语言模型，并通过微调大语言模型融入语义上下文。准确率随 MEG 记录数据量呈对数线性增长。Meta 已以 CC BY-NC 4.0 许可公开 v1 和 v2 的完整训练代码。此为研究发布，在受控实验室环境中对 9 名健康志愿者进行测试；实际应用受限，需要磁屏蔽室和参与者保持静止。
Key points
Brain2Qwerty v2 decodes typed sentences from non-invasive MEG signals without any implant or surgery.
Brain2Qwerty v2 利用非侵入式 MEG 信号解码打字句子，无需植入物或手术。
Average word accuracy is 61% (39% WER), and the best participant reached 78%, compared to only 8% from prior non-invasive methods.
平均词准确率为 61%（词错误率 39%），最佳参与者达到 78%，而此前非侵入式方法仅为 8%。
The architecture combines a convolutional encoder, transformer, character-level language model, and fine-tuned LLMs for semantic correction.
架构结合卷积编码器、Transformer、字符级语言模型和微调的大语言模型进行语义校正。
Accuracy scales log-linearly with data, suggesting that more recordings could further narrow the gap with invasive implants.
准确率随数据量呈对数线性增长，表明更多记录数据可进一步缩小与侵入式植入设备的差距。
Full training code for Brain2Qwerty v1 and v2 is publicly released under CC BY-NC 4.0.
Brain2Qwerty v1 和 v2 的完整训练代码已以 CC BY-NC 4.0 许可公开发布。
Limitations include the need for a magnetically shielded room, a still participant, and testing only on healthy volunteers, not patients with brain injuries.
局限包括需要磁屏蔽室和参与者静止，且仅在健康志愿者而非脑损伤患者上测试。