Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2401.07392 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866913196559826944 |
|---|---|
| author | Scilipoti, Marco Fuster, Marina Ramele, Rodrigo |
| author_facet | Scilipoti, Marco Fuster, Marina Ramele, Rodrigo |
| contents | Deep learning networks have become the de-facto standard in Computer Vision for industry and research. However, recent developments in their cousin, Natural Language Processing (NLP), have shown that there are areas where parameter-less models with strong inductive biases can serve as computationally cheaper and simpler alternatives. We propose such a model for binary image classification: a nearest neighbor classifier combined with a general purpose compressor like Gzip. We test and compare it against popular deep learning networks like Resnet, EfficientNet and Mobilenet and show that it achieves better accuracy and utilizes significantly less space, more than two order of magnitude, within a few-shot setting. As a result, we believe that this underlines the untapped potential of models with stronger inductive biases in few-shot scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_07392 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | A Strong Inductive Bias: Gzip for binary image classification Scilipoti, Marco Fuster, Marina Ramele, Rodrigo Computer Vision and Pattern Recognition Deep learning networks have become the de-facto standard in Computer Vision for industry and research. However, recent developments in their cousin, Natural Language Processing (NLP), have shown that there are areas where parameter-less models with strong inductive biases can serve as computationally cheaper and simpler alternatives. We propose such a model for binary image classification: a nearest neighbor classifier combined with a general purpose compressor like Gzip. We test and compare it against popular deep learning networks like Resnet, EfficientNet and Mobilenet and show that it achieves better accuracy and utilizes significantly less space, more than two order of magnitude, within a few-shot setting. As a result, we believe that this underlines the untapped potential of models with stronger inductive biases in few-shot scenarios. |
| title | A Strong Inductive Bias: Gzip for binary image classification |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2401.07392 |