教育经历
2020-2025年 西安电子科技大学 博士
2016-2020年 西安电子科技大学 本科
研究方向
多模态大模型细粒度视觉理解、空间理解，模态统一范式。
大模型推理加速，模型小型化，模型剪枝。
深度神经网络与人脑神经系统相关机理探究，神经机制建模。
长尾学习，数据异构的联邦学习等。
研究成果
教授课程
人工智能：感知与认知基础
海量数据挖掘
高等代数I
研究成果
2026
[1] Reasoning emerges from constrained inference manifolds in large language models
Nature正刊（已送审），第一作者，2026年，中国人民大学
[2] Compositional Attribute Imbalance in Vision Datasets
发表期刊/会议: AAAI 2026 (AAAI Conference on Artificial Intelligence), pp. 7836-7846
发表时间: 2026年, 第一作者，单位：中国人民大学
[3] ScDiVa: Masked Discrete Diffusion for Joint Modeling of Single-Cell Identity and Expression
发表期刊/会议: ICML 2026 (Forty-Third International Conference on Machine Learning),
发表时间: 2026年，通讯作者，单位：中国人民大学
[4] Federated Manifold Learning (FML): Tackling Domain Heterogeneity with Structural Knowledge Transfer
发表期刊/会议: ICML 2026 (Forty-Third International Conference on Machine Learning),
发表时间: 2026年，通讯作者，单位：中国人民大学
[5] MessToClean: Evidence-Grounded Structure-Preserving Reconstruction for Real-World Degraded Exam Paper Images
发表期刊/会议: ACL 2026 (The 64th Annual Meeting of the Association for Computational Linguistics),
发表时间: 2026年，通讯作者，单位：中国人民大学
[6] FedMC: Federated Manifold Calibration
发表期刊/会议: ICLR 2026 (The Fourteenth International Conference on Learning Representations),
发表时间: 2026年，第一作者，单位：中国人民大学
[7] Calibrating Biased Distribution in VFM-derived Latent Space via Cross-Domain Geometric Consistency
发表期刊/会议: IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI),
发表时间: 2026年，第一作者，单位：中国人民大学
[7] On-Device Large Language Models: A Survey of Model Compression and System Optimization
发表期刊/会议: Artificial Intelligence Review,
发表时间: 2026年，通讯作者，单位：中国人民大学
2025
Geometric Knowledge-Guided Localized Global Distribution Alignment for Federated Learning
Yanbiao Ma, Wei Dai, Wenke Huang, Jiayi Chen
The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR 2025)
Pursuing Better Decision Boundaries for Long-Tailed Object Detection via Category Information Amount
Yanbiao Ma, Wei Dai, Jiayi Chen
The Thirteenth International Conference on Learning Representations (ICLR 2025)
Predicting and Enhancing the Fairness of DNNs with the Curvature of Perceptual Manifolds
Yanbiao Ma, Licheng Jiao, Fang Liu, Maoji Wen, Lingling Li, Wen Ma, Shuyuan Yang, Xu Liu, Puhua Che
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
2024
Unveiling and Mitigating Generalized Biases of DNNs through the Intrinsic Dimensions of Perceptual Manifolds
Yanbiao Ma, Licheng Jiao, Fang Liu, Lingling Li, Wen Ma, Shuyuan Yang, Xu Liu, Puhua Chen
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
Geometric Prior Guided Feature Representation Learning for Long-Tailed Classification
Yanbiao Ma, Licheng Jiao, Fang Liu, Shuyuan Yang, Xu Liu, Puhua Chen
International Journal of Computer Vision (IJCV)
2023
Orthogonal uncertainty representation of data manifold for robust long-tailed learning
Yanbiao Ma, Licheng Jiao, Fang Liu, Shuyuan Yang, Xu Liu, Lingling Li
Proceedings of the 31st ACM International Conference on Multimedia (ACM MM)
Feature distribution representation learning based on knowledge transfer for long-tailed classification
Yanbiao Ma, Licheng Jiao, Fang Liu, Shuyuan Yang, Xu Liu, Puhua Chen
IEEE Transactions on Multimedia
Curvature-balanced feature manifold learning for long-tailed classification
Yanbiao Ma, Licheng Jiao, Fang Liu, Shuyuan Yang, Xu Liu, Lingling Li
The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR 2023)
Delving into semantic scale imbalance
Yanbiao Ma, Licheng Jiao, Fang Liu, Yuxin Li, Shuyuan Yang, Xu Liu
The International Conference on Learning Representations (ICLR2023)
