.png){kind=icon}
Return
Compression-enabled interpretability of voxelwise encoding models
.png){kind=icon}
OA
AI
DOI：10.1371/journal.pcbi.1012822.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Voxelwise encoding models based on convolutional neural networks (CNNs) have emerged as state-of-the-art predictive models of brain activity evoked by natural movies. Despite their superior predictive performance, the huge number of parameters in CNN-based models have made them difficult to interpret. Here, we investigate whether model compression can build more interpretable and more stable CNN-based voxelwise models while maintaining accuracy. We used multiple compression techniques to prune less important CNN filters and connections, a receptive field compression method to select receptive fields with optimal center and size, and principal component analysis to reduce dimensionality. We demonstrate that the model compression improves the accuracy of identifying visual stimuli in a hold-out test set. Additionally, compressed models offer a more stable interpretation of voxelwise pattern selectivity than uncompressed models. Finally, the receptive field-compressed models reveal that the optimal model-based population receptive fields become larger and more centralized along the ventral visual pathway. Overall, our findings support using model compression to build more interpretable voxelwise models.
Keywords:
NATURAL IMAGES
REPRESENTATIONS
Journal
IF:
7.2
Papers:
2.1K
Citations:
3.9W
Organization
Cited Papers
ShareSaveReconstructing Visual Experiences from Brain Activity Evoked by Natural Movies.png)
CURRENT BIOLOGY
OAAI
CURRENT BIOLOGYIF7.5
OAAIShareSaveShareSaveIncreasingly complex representations of natural movies across the dorsal stream are shared between subjectsNeuroImageOAAI
NeuroImageIF0
OAAIShareSaveDeep recurrent neural network reveals a hierarchy of process memory during dynamic natural visionHUMAN BRAIN MAPPINGOAAI
HUMAN BRAIN MAPPINGIF3.3
OAAIShareSaveParametric reverse correlation reveals spatial linearity of retinotopic human V1 BOLD responseNeuroImage
NeuroImageIF0
PREAI
ShareSave
View more