Xinran Li

Cheng Lei, Jie Fan, Xinran Li et al.

Camouflaged Object Segmentation (COS) faces significant challenges due to the scarcity of annotated data, where meticulous pixel-level annotation is both labor-intensive and costly, primarily due to the intricate object-background boundaries. Addressing the core question, "Can COS be effectively achieved in a zero-shot manner without manual annotations for any camouflaged object?" we affirmatively respond and introduce a robust zero-shot COS framework. This framework leverages the inherent local pattern bias of COS and employs a broad semantic feature space derived from salient object segmentation (SOS) for efficient zero-shot transfer. We incorporate an Masked Image Modeling (MIM) based image encoder optimized for Parameter-Efficient Fine-Tuning (PEFT), a Multimodal Large Language Model (M-LLM), and a Multi-scale Fine-grained Alignment (MFA) mechanism. The MIM pre-trained image encoder focuses on capturing essential low-level features, while the M-LLM generates caption embeddings processed alongside these visual cues. These embeddings are precisely aligned using MFA, enabling our framework to accurately interpret and navigate complex semantic contexts. To optimize operational efficiency, we introduce a learnable codebook that represents the M-LLM during inference, significantly reducing computational overhead. Our framework demonstrates its versatility and efficacy through rigorous experimentation, achieving state-of-the-art performance in zero-shot COS with $F_β^w$ scores of 72.9\% on CAMO and 71.7\% on COD10K. By removing the M-LLM during inference, we achieve an inference speed comparable to that of traditional end-to-end models, reaching 18.1 FPS. Code: https://github.com/R-LEI360725/ZSCOS-CaMF

Jiaqi Qiao, Xiujuan Xu, Xinran Li et al.

Multimodal emotion understanding requires effective integration of text, audio, and visual modalities for both discrete emotion recognition and continuous sentiment analysis. We present EGMF, a unified framework combining expert-guided multimodal fusion with large language models. Our approach features three specialized expert networks--a fine-grained local expert for subtle emotional nuances, a semantic correlation expert for cross-modal relationships, and a global context expert for long-range dependencies--adaptively integrated through hierarchical dynamic gating for context-aware feature selection. Enhanced multimodal representations are integrated with LLMs via pseudo token injection and prompt-based conditioning, enabling a single generative framework to handle both classification and regression through natural language generation. We employ LoRA fine-tuning for computational efficiency. Experiments on bilingual benchmarks (MELD, CHERMA, MOSEI, SIMS-V2) demonstrate consistent improvements over state-of-the-art methods, with superior cross-lingual robustness revealing universal patterns in multimodal emotional expressions across English and Chinese. We will release the source code publicly.

Jun Zhang, Xinran Li, Fang-Wei Fu

In this paper, we construct an authentication scheme for multi-receivers and multiple messages based on a linear code $C$. This construction can be regarded as a generalization of the authentication scheme given by Safavi-Naini and Wang. Actually, we notice that the scheme of Safavi-Naini and Wang is constructed with Reed-Solomon codes. The generalization to linear codes has the similar advantages as generalizing Shamir's secret sharing scheme to linear secret sharing sceme based on linear codes. For a fixed message base field $\f$, our scheme allows arbitrarily many receivers to check the integrity of their own messages, while the scheme of Safavi-Naini and Wang has a constraint on the number of verifying receivers $V\leqslant q$. And we introduce access structure in our scheme. Massey characterized the access structure of linear secret sharing scheme by minimal codewords in the dual code whose first component is 1. We slightly modify the definition of minimal codewords in \cite{Massey93}. Let $C$ be a $[V,k]$ linear code. For any coordinate $i\in \{1,2,\cdots,V\}$, a codeword $\vec{c}$ in $C$ is called minimal respect to $i$ if the codeword $\vec{c}$ has component 1 at the $i$-th coordinate and there is no other codeword whose $i$-th component is 1 with support strictly contained in that of $\vec{c}$. Then the security of receiver $R_i$ in our authentication scheme is characterized by the minimal codewords respect to $i$ in the dual code $C^\bot$.