Saved in:
| Main Author: | |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.03297 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912626772017152 |
|---|---|
| author | Gothi, Akshar |
| author_facet | Gothi, Akshar |
| contents | We present a controlled comparison of a convolutional neural network (EfficientNet-B0) and a Vision Transformer (ViT-Base) on SpaceNet under two label-distribution regimes: a naturally imbalanced five-class split and a balanced-resampled split with 700 images per class (70:20:10 train/val/test). With matched preprocessing (224x224, ImageNet normalization), lightweight augmentations, and a 40-epoch budget on a single NVIDIA P100, we report accuracy, macro-F1, balanced accuracy, per-class recall, and deployment metrics (model size and latency). On the imbalanced split, EfficientNet-B0 reaches 93% test accuracy with strong macro-F1 and lower latency; ViT-Base is competitive at 93% with a larger parameter count and runtime. On the balanced split, both models are strong; EfficientNet-B0 reaches 99% while ViT-Base remains competitive, indicating that balancing narrows architecture gaps while CNNs retain an efficiency edge. We release manifests, logs, and per-image predictions to support reproducibility. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_03297 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Convolutional Neural Nets vs Vision Transformers: A SpaceNet Case Study with Balanced vs Imbalanced Regimes Gothi, Akshar Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning We present a controlled comparison of a convolutional neural network (EfficientNet-B0) and a Vision Transformer (ViT-Base) on SpaceNet under two label-distribution regimes: a naturally imbalanced five-class split and a balanced-resampled split with 700 images per class (70:20:10 train/val/test). With matched preprocessing (224x224, ImageNet normalization), lightweight augmentations, and a 40-epoch budget on a single NVIDIA P100, we report accuracy, macro-F1, balanced accuracy, per-class recall, and deployment metrics (model size and latency). On the imbalanced split, EfficientNet-B0 reaches 93% test accuracy with strong macro-F1 and lower latency; ViT-Base is competitive at 93% with a larger parameter count and runtime. On the balanced split, both models are strong; EfficientNet-B0 reaches 99% while ViT-Base remains competitive, indicating that balancing narrows architecture gaps while CNNs retain an efficiency edge. We release manifests, logs, and per-image predictions to support reproducibility. |
| title | Convolutional Neural Nets vs Vision Transformers: A SpaceNet Case Study with Balanced vs Imbalanced Regimes |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2510.03297 |