Batch-CAM: Introduction to better reasoning in convolutional deep learning models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914328173608960 |
|---|---|
| author | Ignesti, Giacomo Moroni, Davide Martinelli, Massimo |
| author_facet | Ignesti, Giacomo Moroni, Davide Martinelli, Massimo |
| contents | Deep learning opacity often impedes deployment in high-stakes domains. We propose a training framework that aligns model focus with class-representative features without requiring pixel-level annotations. To this end, we introduce Batch-CAM, a vectorised implementation of Gradient-weighted Class Activation Mapping that integrates directly into the training loop with minimal computational overhead. We propose two regularisation terms: a Prototype Loss, which aligns individual-sample attention with the global class average, and a Batch-CAM Loss, which enforces consistency within a training batch. These are evaluated using L1, L2, and SSIM metrics. Validated on MNIST and Fashion-MNIST using ResNet18 and ConvNeXt-V2, our method generates significantly more coherent and human-interpretable saliency maps compared to baselines. While maintaining competitive classification accuracy, the framework successfully suppresses spurious feature activation, as evidenced by qualitative reconstruction analysis. Batch-CAM appears to offer a scalable pathway for training intrinsically interpretable models by leveraging batch-level statistics to guide feature extraction, effectively bridging the gap between predictive performance and explainability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_00664 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Batch-CAM: Introduction to better reasoning in convolutional deep learning models Ignesti, Giacomo Moroni, Davide Martinelli, Massimo Artificial Intelligence Computer Vision and Pattern Recognition 68 I.2; I.4 Deep learning opacity often impedes deployment in high-stakes domains. We propose a training framework that aligns model focus with class-representative features without requiring pixel-level annotations. To this end, we introduce Batch-CAM, a vectorised implementation of Gradient-weighted Class Activation Mapping that integrates directly into the training loop with minimal computational overhead. We propose two regularisation terms: a Prototype Loss, which aligns individual-sample attention with the global class average, and a Batch-CAM Loss, which enforces consistency within a training batch. These are evaluated using L1, L2, and SSIM metrics. Validated on MNIST and Fashion-MNIST using ResNet18 and ConvNeXt-V2, our method generates significantly more coherent and human-interpretable saliency maps compared to baselines. While maintaining competitive classification accuracy, the framework successfully suppresses spurious feature activation, as evidenced by qualitative reconstruction analysis. Batch-CAM appears to offer a scalable pathway for training intrinsically interpretable models by leveraging batch-level statistics to guide feature extraction, effectively bridging the gap between predictive performance and explainability. |
| title | Batch-CAM: Introduction to better reasoning in convolutional deep learning models |
| topic | Artificial Intelligence Computer Vision and Pattern Recognition 68 I.2; I.4 |
| url | https://arxiv.org/abs/2510.00664 |