Nip Rumors in the Bud: Retrieval-Guided Topic-Level Adaptation for Test-Time Fake News Video Detection

Jian Lang, Rongpei Hong, Ting Zhong, Yong Wang, Fan Zhou†

KDD 2026 | CCF A | PDF | Github

• RADAR, the first work to achieves the test-time adaptation of the Fake News Video Detection,
• Enabling fast adaptation to evolving news videos with shifting topic-level distributions in the dynamic world.

Shedding the Facades, Connecting the Domains: Detecting Shifting Multimodal Hate Video with Test-Time Adaptation

Jiao Li, Jian Lang, Xikai Tang†, Ting Zhong, Fan Zhou

AAAI 2026 | CCF A | PDF | Github

• SCANNER, the first test-time adaptation framework tailored for distribution shifting hate video detection.

REDEEMing Modality Information Loss: Retrieval-Guided Conditional Generation for Severely Modality Missing Learning

Jian Lang, Rongpei Hong, Zhangtao Cheng, Ting Zhong, Fan Zhou†

KDD 2025 | CCF A | PDF | Github

• REDEEM, the extension work of our RAGPT.
• Proposing a retrieval-guided conditional generation paradigm for enhancing the modality-missing robustness of pre-trained Multimodal Transformer.

Retrieval-Augmented Dynamic Prompt Tuning for Incomplete Multimodal Learning

Jian Lang*, Zhangtao Cheng*, Ting Zhong, Fan Zhou†

AAAI 2025 | CCF A | PDF | Github |

• RAGPT, a novel retrieval-augmented dynamic prompt-tuning framework for enhancing the modality-missing robustness of pre-trained Multimodal Transformer.

From Shallow Humor to Metaphor: Towards Label-Free Harmful Meme Detection via LMM Agent Self-Improvement

Jian Lang, Rongpei Hong, Ting Zhong, Leiting Chen, Qiang Gao, Fan Zhou†

KDD 2026 | CCF A | PDF | Github

• ALARM, the first label-free harmful meme detection framework powered by LMM self-improvement
• Enabling prompt and robust adaptation to evolving topics and themes of harmful web memes.

Borrowing Eyes for the Blind Spot: Overcoming Data Scarcity in Malicious Video Detection via Cross-Domain Retrieval Augmentation

Rongpei Hong*, Jian Lang*, Ting Zhong, Fan Zhou†

ICCV 2025 | CCF A | PDF | Github

• CRAVE, a novel cross-domain retrieval augmentation framework that transfers knowledge from resource-rich image-text domain to enhance malicious video detection with scarce training data.

Biting Off More Than You Can Detect: Retrieval-Augmented Multimodal Experts for Short Video Hate Detection

Jian Lang, Rongpei Hong, Jin Xu, Xovee Xu, Yili Li, Fan Zhou†

WWW 2025 | CCF A | PDF | Github

• MoRE, a novel mixture of retrieval-augmented multimodal experts framework to enhance hate video detection.

Following Clues, Approaching the Truth: Explainable Micro-Video Rumor Detection via Chain-of-Thought Reasoning

Rongpei Hong, Jian Lang, Jin Xu, Zhangtao Cheng, Ting Zhong†, Fan Zhou

WWW 2025 | CCF A | PDF | Github

• ExMRD, the first explainable fake news video detection framework powered by the Chain-of-Thought Reasoning.

REAL: Retrieval-Augmented Prototype Alignment for Improved Fake News Video Detection

Yili Li, Jian Lang, Rongpei Hong, Qing Chen, Zhangtao Cheng, Jia Chen, Ting Zhong, Fan Zhou†

ICME 2025 | CCF B | PDF | Github

• REAL, a novel model-agnostic framework that generates manipulation-aware representations to enhance existing methods in detecting fake news videos with only subtle modifications to the original authentic ones.

Predicting Micro-video Popularity via Multi-modal Retrieval Augmentation

Ting Zhong, Jian Lang, Yifan Zhang, Zhangtao Cheng, Kunpeng Zhang, Fan Zhou†

SIGIR 2024 | CCF A | PDF | Github |

• MMRA, a multi-modal retrieval-augmented popularity prediction model that enhances prediction accuracy using relevant retrieved information.

TAMEing Long Contexts in Personalization: Towards Training-Free and State-Aware MLLM Personalized Assistant

Rongpei Hong, Jian Lang, Ting Zhong†, Yong Wang, Fan Zhou

KDD 2026 | CCF A | PDF | Github

• TAME, the first training-free and state-aware personalized Multimodal Large Multimodal Model assistant powered by double memories.