Xiaojun Ye

Xiaojun Ye

Auto-generated profile, based on publicly available information

Xiaojun Ye

Zhejiang University, Hangzhou, China

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 December 2024
Verified
Conference Paper
Open access
YSYiren Song
Yiren Song
SHShijie Huang
Shijie Huang
CYChen Yao
Chen Yao
HCHai Ci
Hai Ci
XYXiaojun Ye
Xiaojun Ye
MSMike Zheng Shou
Mike Zheng Shou


1 September 2025
Verified
Conference Paper
Open access
PDF file
XYXiaojun Ye
Xiaojun Ye
GLGuanbao Liang
Guanbao Liang
CWChun Wang
Chun Wang
LLLiangcheng Li
Liangcheng Li
PKPengfei Ke
Pengfei Ke
SZSheng Zhou
Sheng Zhou
The increasing demands in analyzing complex associated scenes pose necessities to researching multi-image understanding abilities. Compared with understanding individual images, both the alignments and differences between images are essential aspects of understanding the intricate relationships for multi-image inference tasks. However, existing benchmarks face difficulties in addressing both of these aspects simultaneously, resulting in obstacles to modeling relationships under various granularities and domains of images. In this paper, we introduce M4Bench to enhance the capability of aligning and distinguishing multi-images with multi-domain multi-granularity comparison. We carefully design five comparison tasks related to coarse and fine-grained granularities in single and multiple domains of images and evaluate them on 13 state-of-the-art multi-modal large language models with various sizes. Besides, we analyze the evaluation results and provide several observations and viewpoints for the multi-image understanding research. The data and evaluation code are available at https://github.com/eaglelab-zju/M4Bench.

26 May 2025 in arXiv
Verified
Preprint
Open access
PDF file
XHXinmiao Hu
Xinmiao Hu
CWChun Wang
Chun Wang
RARuihe An
Ruihe An
CSChenYu Shao
ChenYu Shao
XYXiaojun Ye
Xiaojun Ye
LLLiangcheng Li
Liangcheng Li
Multimodal Large Language Models (MLLMs) have demonstrated strong performance in visual understanding tasks, yet they often suffer from object hallucinations--generating descriptions of objects that are inconsistent with or entirely absent from the input. This issue is closely related to dataset biases, where frequent co-occurrences of objects lead to entangled semantic representations across modalities. As a result, models may erroneously activate object representations that are commonly associated with the input but not actually present. To address this, we propose a causality-driven disentanglement framework that mitigates hallucinations through causal intervention. Our approach includes a Causal-Driven Projector in the visual pathway and a Causal Intervention Module integrated into the final transformer layer of the language model. These components work together to reduce spurious correlations caused by biased training data. Experimental results show that our method significantly reduces hallucinations while maintaining strong performance on multiple multimodal benchmarks. Visualization analyses further confirm improved separability of object representations. The code is available at: https://github.com/IgniSavium/Causal-LLaVA

9 November 2026
Verified
Conference Paper
Open access
XYXiaojun Ye
Xiaojun Ye
CWChun Wang
Chun Wang
JCJunliang Chen
Junliang Chen
SZSheng Zhou
Sheng Zhou
RWRui Wang
Rui Wang
JBJiajun Bu
Jiajun Bu