Dataset Condensation with Distribution Matching
This work addresses the problem of expensive dataset condensation for machine learning practitioners, offering an efficient solution that is incremental but effective for larger datasets and applications like continual learning.
The authors tackled the high computational cost of dataset condensation by proposing a method that matches feature distributions between synthetic and original images in sampled embedding spaces, achieving comparable or better performance while significantly reducing synthesis costs.
Computational cost of training state-of-the-art deep models in many learning problems is rapidly increasing due to more sophisticated models and larger datasets. A recent promising direction for reducing training cost is dataset condensation that aims to replace the original large training set with a significantly smaller learned synthetic set while preserving the original information. While training deep models on the small set of condensed images can be extremely fast, their synthesis remains computationally expensive due to the complex bi-level optimization and second-order derivative computation. In this work, we propose a simple yet effective method that synthesizes condensed images by matching feature distributions of the synthetic and original training images in many sampled embedding spaces. Our method significantly reduces the synthesis cost while achieving comparable or better performance. Thanks to its efficiency, we apply our method to more realistic and larger datasets with sophisticated neural architectures and obtain a significant performance boost. We also show promising practical benefits of our method in continual learning and neural architecture search.