ERank in Latent Space as an Image-Complexity and Richness Measure
Not on the map. Why?
Abstract (arXiv)
We propose the effective rank (ERank) of the channel covariance of an image's deep feature map as a per-sample, label-free measure of visual richness, computed from a single forward pass through a frozen pretrained encoder. ERank counts how many decorrelated channel directions an image activates, and we characterize its properties, including its behavior under noise. Empirically, ERank orders images from plain to visually rich, correlates with codec bitrate, sharpness, and edge density, and correlates with human complexity annotations on IC9600 with r = 0.72. As a data-selection criterion, removing low-ERank samples improves super-resolution and removing high-ERank samples improves OCR, in both pretraining and finetuning, while selection does not help classification, segmentation, or denoising. ERank is thus a cheap richness signal, useful exactly when task difficulty is governed by input richness.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 30 of its 55 references that are in the map. First posted on arXiv 21 Jul 2026.
Builds on
30 of 55 references.
References · 30
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
- 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
- 2016Improved Techniques for Training GANscited by 2,970
- 2017Rethinking Atrous Convolution for Semantic Image Segmentationcited by 2,161
- 2017Enhanced Deep Residual Networks for Single Image Super-Resolutioncited by 1,552
- 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
- 2020Prevalence of Neural Collapse during the terminal phase of deep learning trainingcited by 719
- 2018An Empirical Study of Example Forgetting during Deep Neural Network Learningcited by 678
- 2021Deep Learning on a Data Diet: Finding Important Examples Early in Trainingcited by 533
- 2022Beyond neural scaling laws: beating power law scaling via data pruningcited by 528
- 2015An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognitioncited by 520
18 more
- 2020Reliable Fidelity and Diversity Metrics for Generative Modelscited by 434
- 2021Dynamic Neural Networks: A Surveycited by 408
- 2018Assessing Generative Models via Precision and Recallcited by 406
- 2019Implicit Regularization in Deep Matrix Factorizationcited by 405
- 2021Understanding Dimensional Collapse in Contrastive Self-supervised Learningcited by 378
- 2017Implicit Regularization in Matrix Factorizationcited by 322
- 2021The Intrinsic Dimension of Images and Its Impact on Learningcited by 315
- 2022The Vendi Score: A Diversity Evaluation Metric for Machine Learningcited by 311
- 2023SemDeDup: Data-efficient learning at web-scale through semantic deduplicationcited by 303
- 2019Intrinsic dimension of data representations in deep neural networkscited by 282
- 2021The Low-Rank Simplicity Bias in Deep Networkscited by 154
- 2022RankMe: Assessing the downstream performance of pretrained self-supervised representations by their rankcited by 153
- 2022DeepCore: A Comprehensive Library for Coreset Selection in Deep Learningcited by 147
- 2021Not All Images are Worth 16x16 Words: Dynamic Transformers for Efficient Image Recognitioncited by 144
- 2025A Coreset Selection of Coreset Selection Literature: Introduction and Recent Advancescited by 28
- 2023Repeated Random Sampling for Minimizing the Time-to-Accuracy of Learningcited by 26
- 2023Data pruning and neural scaling laws: fundamental limitations of score-based algorithmscited by 11
- 2025Image Complexity-Aware Adaptive Retrieval for Efficient Vision-Language Modelscited by 1
Built on by
Not cited yet.