Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.05599 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914307316383744 |
|---|---|
| author | Sawada, Takito Iwata, Akinori Okuda, Masahiro |
| author_facet | Sawada, Takito Iwata, Akinori Okuda, Masahiro |
| contents | Convolutional Neural Networks (CNNs) exhibit a well-known texture bias, prioritizing local patterns over global shapes - a tendency inherent to their convolutional architecture. While this bias is beneficial for texture-rich natural images, it often degrades performance on shape-dominant data such as illustrations and sketches. Although prior work has proposed shape-biased models to mitigate this issue, these approaches lack a quantitative metric for identifying which datasets would actually benefit from such modifications. To address this limitation, we propose a data-driven metric that quantifies the shape-texture balance within a dataset by computing the Structural Similarity Index (SSIM) between an image's luminance (Y) channel and its L0-smoothed counterpart. Building on this metric, we introduce a computationally efficient adaptation method that promotes shape bias by modifying the dilation of max-pooling operations while keeping convolutional weights frozen. Experimental results demonstrate consistent accuracy improvements on shape-dominant datasets, particularly in low-data regimes where full fine-tuning is impractical, requiring training only the final classification layer. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_05599 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Quantifying and Inducing Shape Bias in CNNs via Max-Pool Dilation Sawada, Takito Iwata, Akinori Okuda, Masahiro Computer Vision and Pattern Recognition Machine Learning I.2.10; I.4.7 Convolutional Neural Networks (CNNs) exhibit a well-known texture bias, prioritizing local patterns over global shapes - a tendency inherent to their convolutional architecture. While this bias is beneficial for texture-rich natural images, it often degrades performance on shape-dominant data such as illustrations and sketches. Although prior work has proposed shape-biased models to mitigate this issue, these approaches lack a quantitative metric for identifying which datasets would actually benefit from such modifications. To address this limitation, we propose a data-driven metric that quantifies the shape-texture balance within a dataset by computing the Structural Similarity Index (SSIM) between an image's luminance (Y) channel and its L0-smoothed counterpart. Building on this metric, we introduce a computationally efficient adaptation method that promotes shape bias by modifying the dilation of max-pooling operations while keeping convolutional weights frozen. Experimental results demonstrate consistent accuracy improvements on shape-dominant datasets, particularly in low-data regimes where full fine-tuning is impractical, requiring training only the final classification layer. |
| title | Quantifying and Inducing Shape Bias in CNNs via Max-Pool Dilation |
| topic | Computer Vision and Pattern Recognition Machine Learning I.2.10; I.4.7 |
| url | https://arxiv.org/abs/2601.05599 |