影响力指数

90.46/100

前 0.5%

全站排名 #346

发表论文20 篇

平均评分6.9

年均产出6.7 篇/年

Xi Peng

Full Professor@Sichuan University·中国·OpenReview

研究方向

image enhancement · multi-view · multi-modal · noisy label · noisy correspondence · misalignment · mismatch · nas · contrastive learning · visual transformer · dehazing · super resolution · deblur · image recovery · denoise · clustering · unsupervised representation learning · sparse representation · low rank representation

7.5

DOBB: Decoupled Object-level Bridging with von Mises–Fisher Uncertainty for Hallucination Detection in MLLMs

ICLR 2026Withdrawn

7.8

Robust Cross-modal Alignment Learning for Cross-Scene Spatial Reasoning and Grounding

NeurIPS 2025Poster

7.8

Neighbor-aware Contrastive Disambiguation for Cross-Modal Hashing with Redundant Annotations

NeurIPS 2025Spotlight

7.8

Deep Unsupervised Hashing via External Guidance

ICML 2025Poster

7.8

Deep Fuzzy Multi-view Learning for Reliable Classification

ICML 2025Poster

7.8

LLaVA-ReID: Selective Multi-image Questioner for Interactive Person Re-Identification

ICML 2025Poster

通讯

7.5

Test-time Adaptation for Cross-modal Retrieval with Query Shift

ICLR 2025Spotlight

7.3

Learning Source-Free Domain Adaptation for Visible-Infrared Person Re-Identification

NeurIPS 2025Poster

6.8

Conditional Representation Learning for Customized Tasks

NeurIPS 2025Spotlight

通讯

5.5

Visual Abstraction: A Plug-and-Play Approach for Text-Visual Retrieval

合作者 (20)

Xi Peng

Learning with Dual-level Noisy Correspondence for Multi-modal Entity Alignment

Uncover Underlying Correspondence for Robust Multi-view Clustering

AlignPrune: Robust Dynamic Data Pruning through Loss Trajectory Alignment

DOBB: Decoupled Object-level Bridging with von Mises–Fisher Uncertainty for Hallucination Detection in MLLMs

Robust Cross-modal Alignment Learning for Cross-Scene Spatial Reasoning and Grounding

Neighbor-aware Contrastive Disambiguation for Cross-Modal Hashing with Redundant Annotations

Deep Unsupervised Hashing via External Guidance

Deep Fuzzy Multi-view Learning for Reliable Classification

LLaVA-ReID: Selective Multi-image Questioner for Interactive Person Re-Identification

Test-time Adaptation for Cross-modal Retrieval with Query Shift

Learning Source-Free Domain Adaptation for Visible-Infrared Person Re-Identification

Conditional Representation Learning for Customized Tasks

Visual Abstraction: A Plug-and-Play Approach for Text-Visual Retrieval