Xinyue Bi — accepted AI-conference papers

2026

Hard Labels In! Rethinking the Role of Hard Labels in Mitigating Local Semantic Drift

Jiacheng Cui, Bingkui Tong, Xinyue Bi, Xiaohan Zhao, Jiacheng Liu, Zhiqiang Shen

ICML 2026poster

Soft labels from teacher models are a $\textit{de facto}$ practice for knowledge transfer and large-scale dataset distillation (e.g., SRe$^2$L, RDED, LPLD). However, when we limit the number of crops per image to reduce the substantial cost of storing precomputed soft labels, these methods suffer se…

Cited by 0 Source ↗Scholar ↗

2025

FADRM: Fast and Accurate Data Residual Matching for Dataset Distillation

Jiacheng Cui, Xinyue Bi, Yaxin Luo, Xiaohan Zhao, Jiacheng Liu, Zhiqiang Shen

NeurIPS 2025poster

Residual connection has been extensively studied and widely applied at the model architecture level. However, its potential in the more challenging data-centric approaches remains unexplored. In this work, we introduce the concept of ***Data Residual Matching*** for the first time, leveraging data-l…

Cited by 0 Source ↗code ↗Scholar ↗