Distilling Long-tailed Datasets

• URL: http://arxiv.org/abs/2408.14506v1
• Date: Sat, 24 Aug 2024 15:36:36 GMT
• Title: Distilling Long-tailed Datasets
• Authors: Zhenghao Zhao, Haoxuan Wang, Yuzhang Shang, Kai Wang, Yan Yan,
• Abstract summary: We propose a novel long-tailed dataset distillation method, Long-tailed dataset Aware distillation (LAD) LAD reduces the distance between the student and the biased expert trajectories and prevents the tail class bias from being distilled to the synthetic dataset. This work pioneers the field of long-tailed dataset distillation (LTDD), marking the first effective effort to distill long-tailed datasets.
• Score: 13.330572317331198
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Dataset distillation (DD) aims to distill a small, information-rich dataset from a larger one for efficient neural network training. However, existing DD methods struggle with long-tailed datasets, which are prevalent in real-world scenarios. By investigating the reasons behind this unexpected result, we identified two main causes: 1) Expert networks trained on imbalanced data develop biased gradients, leading to the synthesis of similarly imbalanced distilled datasets. Parameter matching, a common technique in DD, involves aligning the learning parameters of the distilled dataset with that of the original dataset. However, in the context of long-tailed datasets, matching biased experts leads to inheriting the imbalance present in the original data, causing the distilled dataset to inadequately represent tail classes. 2) The experts trained on such datasets perform suboptimally on tail classes, resulting in misguided distillation supervision and poor-quality soft-label initialization. To address these issues, we propose a novel long-tailed dataset distillation method, Long-tailed Aware Dataset distillation (LAD). Specifically, we propose Weight Mismatch Avoidance to avoid directly matching the biased expert trajectories. It reduces the distance between the student and the biased expert trajectories and prevents the tail class bias from being distilled to the synthetic dataset. Moreover, we propose Adaptive Decoupled Matching, which jointly matches the decoupled backbone and classifier to improve the tail class performance and initialize reliable soft labels. This work pioneers the field of long-tailed dataset distillation (LTDD), marking the first effective effort to distill long-tailed datasets.

Related papers

• Exploring the Impact of Dataset Bias on Dataset Distillation [10.742404631413029]
  We investigate the influence of dataset bias on Dataset Distillation (DD) DD is a technique to synthesize a smaller dataset that preserves essential information from the original dataset. Experiments demonstrate that biases present in the original dataset significantly impact the performance of the synthetic dataset.
  arXiv  Detail & Related papers  (2024-03-24T06:10:22Z)
• Importance-Aware Adaptive Dataset Distillation [53.79746115426363]
  Development of deep learning models is enabled by the availability of large-scale datasets. dataset distillation aims to synthesize a compact dataset that retains the essential information from the large original dataset. We propose an importance-aware adaptive dataset distillation (IADD) method that can improve distillation performance.
  arXiv  Detail & Related papers  (2024-01-29T03:29:39Z)
• Embarassingly Simple Dataset Distillation [0.0]
  We tackle dataset distillation at its core by treating it directly as a bilevel optimization problem. A deeper dive into the nature of distilled data unveils pronounced intercorrelation. We devise a boosting mechanism that generates distilled datasets that contain subsets with near optimal performance across different data budgets.
  arXiv  Detail & Related papers  (2023-11-13T02:14:54Z)
• AST: Effective Dataset Distillation through Alignment with Smooth and High-Quality Expert Trajectories [18.266786462036553]
  We propose an effective DD framework named AST, standing for Alignment with Smooth and high-quality expert Trajectories. We conduct extensive experiments on datasets of different scales, sizes, and resolutions.
  arXiv  Detail & Related papers  (2023-10-16T16:13:53Z)
• Distill Gold from Massive Ores: Bi-level Data Pruning towards Efficient Dataset Distillation [96.92250565207017]
  We study the data efficiency and selection for the dataset distillation task. By re-formulating the dynamics of distillation, we provide insight into the inherent redundancy in the real dataset. We find the most contributing samples based on their causal effects on the distillation.
  arXiv  Detail & Related papers  (2023-05-28T06:53:41Z)
• Dataset Distillation: A Comprehensive Review [76.26276286545284]
  dataset distillation (DD) aims to derive a much smaller dataset containing synthetic samples, based on which the trained models yield performance comparable with those trained on the original dataset. This paper gives a comprehensive review and summary of recent advances in DD and its application.
  arXiv  Detail & Related papers  (2023-01-17T17:03:28Z)
• A Comprehensive Survey of Dataset Distillation [73.15482472726555]
  It has become challenging to handle the unlimited growth of data with limited computing power. Deep learning technology has developed unprecedentedly in the last decade. This paper provides a holistic understanding of dataset distillation from multiple aspects.
  arXiv  Detail & Related papers  (2023-01-13T15:11:38Z)
• Dataset Distillation by Matching Training Trajectories [75.9031209877651]
  We propose a new formulation that optimize our distilled data to guide networks to a similar state as those trained on real data. Given a network, we train it for several iterations on our distilled data and optimize the distilled data with respect to the distance between the synthetically trained parameters and the parameters trained on real data. Our method handily outperforms existing methods and also allows us to distill higher-resolution visual data.
  arXiv  Detail & Related papers  (2022-03-22T17:58:59Z)
• LiDAR dataset distillation within bayesian active learning framework: Understanding the effect of data augmentation [63.20765930558542]
  Active learning (AL) has re-gained attention recently to address reduction of annotation costs and dataset size. This paper performs a principled evaluation of AL based dataset distillation on (1/4th) of the large Semantic-KITTI dataset. We observe that data augmentation achieves full dataset accuracy using only 60% of samples from the selected dataset configuration.
  arXiv  Detail & Related papers  (2022-02-06T00:04:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.