Leveraging Recurrent Neural Networks for Predicting Motor Movements from Primate Motor Cortex Neural Recordings
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910680080187392 |
|---|---|
| author | Wang, Yuanxi Wang, Zuowen Liu, Shih-Chii |
| author_facet | Wang, Yuanxi Wang, Zuowen Liu, Shih-Chii |
| contents | This paper presents an efficient deep learning solution for decoding motor movements from neural recordings in non-human primates. An Autoencoder Gated Recurrent Unit (AEGRU) model was adopted as the model architecture for this task. The autoencoder is only used during the training stage to achieve better generalization. Together with the preprocessing techniques, our model achieved 0.71 $R^2$ score, surpassing the baseline models in Neurobench and is ranked first for $R^2$ in the IEEE BioCAS 2024 Grand Challenge on Neural Decoding. Model pruning is also applied leading to a reduction of 41.4% of the multiply-accumulate (MAC) operations with little change in the $R^2$ score compared to the unpruned model. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_22283 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Leveraging Recurrent Neural Networks for Predicting Motor Movements from Primate Motor Cortex Neural Recordings Wang, Yuanxi Wang, Zuowen Liu, Shih-Chii Signal Processing Machine Learning Neurons and Cognition This paper presents an efficient deep learning solution for decoding motor movements from neural recordings in non-human primates. An Autoencoder Gated Recurrent Unit (AEGRU) model was adopted as the model architecture for this task. The autoencoder is only used during the training stage to achieve better generalization. Together with the preprocessing techniques, our model achieved 0.71 $R^2$ score, surpassing the baseline models in Neurobench and is ranked first for $R^2$ in the IEEE BioCAS 2024 Grand Challenge on Neural Decoding. Model pruning is also applied leading to a reduction of 41.4% of the multiply-accumulate (MAC) operations with little change in the $R^2$ score compared to the unpruned model. |
| title | Leveraging Recurrent Neural Networks for Predicting Motor Movements from Primate Motor Cortex Neural Recordings |
| topic | Signal Processing Machine Learning Neurons and Cognition |
| url | https://arxiv.org/abs/2410.22283 |