# Learning scale-variant and scale-invariant features for deep image classification

arXiv: https://arxiv.org/abs/1602.01255 · HTML: https://mlsanity.com/paper/1602.01255/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Feb 2016.

## Built on by (most cited first, 20 of 20)

- [Understanding and Creating Art with AI: Review and Outlook](https://mlsanity.com/paper/2102.09109/) (2021)
- [Scale-invariant scale-channel networks: Deep networks that generalise to previously unseen scales](https://mlsanity.com/paper/2106.06418/) (2021)
- [KonX: Cross-Resolution Image Quality Assessment](https://mlsanity.com/paper/2212.05813/) (2022)
- [Attention Mechanisms in Medical Image Segmentation: A Survey](https://mlsanity.com/paper/2305.17937/) (2023)
- [Knee Osteoarthritis Severity Prediction using an Attentive Multi-Scale Deep Convolutional Neural Network](https://mlsanity.com/paper/2106.14292/) (2021)
- [Survey of Automatic Plankton Image Recognition: Challenges, Existing Solutions and Future Perspectives](https://mlsanity.com/paper/2305.11739/) (2023)
- [Art Authentication with Vision Transformers](https://mlsanity.com/paper/2307.03039/) (2023)
- [Semi-Supervised Object Detection: A Survey on Progress from CNN to Transformer](https://mlsanity.com/paper/2407.08460/) (2024)
- [CNN-based Ego-Motion Estimation for Fast MAV Maneuvers](https://mlsanity.com/paper/2101.01841/) (2021)
- [Just a Matter of Scale? Reevaluating Scale Equivariance in Convolutional Neural Networks](https://mlsanity.com/paper/2211.10288/) (2022)
- [Scale-Preserving Automatic Concept Extraction (SPACE)](https://mlsanity.com/paper/2308.06022/) (2023)
- [Scale Invariant Domain Generalization Image Recapture Detection](https://mlsanity.com/paper/2110.03496/) (2021)
- [D2DF2WOD: Learning Object Proposals for Weakly-Supervised Object Detection via Progressive Domain Adaptation](https://mlsanity.com/paper/2212.01376/) (2022)
- [IrrNet: Advancing Irrigation Mapping with Incremental Patch Size Training on Remote Sensing Imagery](https://mlsanity.com/paper/2404.11762/) (2024)
- [Art Beyond Semantics: Sheaf-Informed Contrastive Learning for Multi-Relational Representations](https://mlsanity.com/paper/2607.16321/) (2026)
- [The Untapped Potential of Off-the-Shelf Convolutional Neural Networks](https://mlsanity.com/paper/2103.09891/) (2021)
- [Can Super Resolution be used to improve Human Pose Estimation in Low Resolution Scenarios?](https://mlsanity.com/paper/2107.02108/) (2021)
- [Deep learning-based denoising for fast time-resolved flame emission spectroscopy in high-pressure combustion environment](https://mlsanity.com/paper/2208.12544/) (2022)
- [A study on the invariance in security whatever the dimension of images for the steganalysis by deep-learning](https://mlsanity.com/paper/2302.11527/) (2023)
- [Geometric Mean Pooling for Equal-Weight Multiplicative Coarse-Graining](https://mlsanity.com/paper/2609.21876/) (2026)

## Abstract (arXiv)

Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks. The variation in image resolutions, sizes of objects and patterns depicted, and image scales, hampers CNN training and performance, because the task-relevant information varies over spatial scales. Previous work attempting to deal with such scale variations focused on encouraging scale-invariant CNN representations. However, scale-invariant representations are incomplete representations of images, because images contain scale-variant information as well. This paper addresses the combined development of scale-invariant and scale-variant representations. We propose a multi- scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple scales. The results show that our multi-scale CNN outperforms single-scale CNN. This leads to the conclusion that encouraging the combined development of a scale-invariant and scale-variant representation in CNNs is beneficial to image recognition performance.