Jinjin Chi, Taoping Liu, Mengtao Yin, Ximing Li, Yongcheng Jing, Dacheng Tao
The paper introduces a flow matching-based framework for disentangled representation learning that improves semantic alignment and disentanglement scores by using a non-overlap regularizer to reduce factor interference.
This research explores a new way to understand and represent complex data by breaking it down into simpler, independent factors. The authors propose a method that uses flow matching to create these representations, which helps in understanding the data more clearly. They also introduce a technique to ensure that these factors do not interfere with each other, making the representations more accurate and meaningful. The approach shows promise, outperforming existing methods in terms of clarity and control over the data representation.
