Author name cluster

Jiaming Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

13 papers

2 author rows

AAAI Conference 2026 Conference Paper

Active Multi-source Domain Adaptation for Multimodal Fake News Detection

Yanping Chen
Weijie Shi
Mengze Li
Yue Cui
Jiaming Li
Ruiyuan Zhang
Hao Chen
Hanghui Guo

Multimodal fake news detection plays a crucial role in combating online misinformation. The inherent domain diversity of news in the real world has driven the development of cross-domain detection methods. However, these detection methods either suffer from significant performance degradation due to semantic and deception pattern shifts between the training (source) and test (target) domains or heavily rely on annotated labels. To address the problems, we propose ADOSE, an active multi-source domain adaptation framework for multimodal fake news detection which actively annotates a small subset of target samples to improve detection performance. Specifically, for domain shifts, we design a multi-expert classifier network based on refined features to comprehensively capture and adapt to the semantic space and deception patterns of news across different domains. To maximize adaptation performance with limited annotation cost, we propose a least-disagree uncertainty selector equipped with a diversity calculator for selecting the most informative samples. The selector leverages the uncertainty of inconsistent predictions before and after perturbations by multiple classifiers as an indicator of unfamiliar samples. It further incorporates diversity scores derived from multi-view features to ensure the chosen samples achieve maximal coverage of target domain features. The extensive experiments on multiple datasets show that ADOSE outperforms existing domain adaptation methods by 2.45% ~ 9.1%, indicating the superiority of our model.

PDF Details DOI

AAAI Conference 2026 Conference Paper

TIV: Thought Injection via Vectors for Efficient Reasoning in Large Reasoning Models

Yi Cao
Weijie Shi
Wei-Jie Xu
Yucheng Shen
Yue Cui
Hanghui Guo
Shimin Di
Ziyi Liu

Large Reasoning Models (LRMs) have recently demonstrated impressive performance across a range of reasoning tasks by generating intermediate thoughts. However, these models can suffer from overthinking—generating excessive tokens that contribute little to final accuracy while increasing inference cost. To mitigate this, we propose TIV (Thought Injection via Vectors), an innovative framework that compresses token-level reasoning into compact vectors without sacrificing performance. Rather than generating explicit thoughts, TIV injects learnable vectors into the post-attention hidden states of the final token across Transformer layers, enabling implicit and lightweight reasoning. We further introduce a two-stage reinforcement learning strategy: the first stage calibrates the model's reasoning distribution, and the second distills it into a vector-based policy optimized for both accuracy and brevity. Experiments on three reasoning benchmarks show that TIV preserves over 99% of the original accuracy while reducing output length by more than 65% on average, reaching up to 80% in some cases. Moreover, TIV consistently achieves superior trade-offs between accuracy and efficiency compared to existing methods, distinguishing itself as a state-of-the-art (SOTA) approach for efficient reasoning in LRMs.

PDF Details DOI

AAAI Conference 2025 Conference Paper

A Multi-Focus-Driven Multi-Branch Network for Robust Multimodal Sentiment Analysis

Chuanqi Tao
Jiaming Li
Tianzi Zang
Peng Gao

Multimodal sentiment analysis aims to integrate diverse modalities for precise emotional interpretation. However, external factors such as sensor malfunctions or network issues may disrupt certain modalities. This may lead to missing data, which poses challenges in real-world deployment. Most existing approaches focus on designing feature reconstruction strategies, overlooking the collaborative integration of reconstruction and fusion strategies. Moreover, they fail to capture the relationships between features in the global dimension and those in the local dimension. These limitations hinder the full capture of the complex nature of multimodal data, especially in scenarios involving missing modalities. To address the above issues, this paper proposes a robust model named MFMB-Net with multiple branches for feature multi-focus fusion and reconstruction. We design a two-stream fusion branch where macro-fusion focuses on the fusion of features in the global dimension and micro-fusion targets local dimension features. This dual-stream fusion branch distributes multi-focus across both pathways, simultaneously capturing global coarse-grained and local fine-grained features. Additionally, the reconstruction branch interacts collaboratively with the fusion branch to reconstruct and enhance the missing data. It integrates the reconstructed feature information with the fused information thus refining the representation fidelity of the missing information. Experiments performed on two benchmarks show that our approach obtains results superior to state-of-the-art models.

PDF Details DOI

ICML Conference 2025 Conference Paper

BSemiFL: Semi-supervised Federated Learning via a Bayesian Approach

Haozhao Wang
Shengyu Wang
Jiaming Li
Hao Ren 0001
Xingshuo Han
Wenchao Xu 0001
Shangwei Guo
Tianwei Zhang 0004

Semi-supervised Federated Learning (SSFL) is a promising approach that allows clients to collaboratively train a global model in the absence of their local data labels. The key step of SSFL is the re-labeling where each client adopts two types of available models, namely global and local models, to re-label the local data. While various technologies such as using the global model or the average of two models have been proposed to conduct the re-labeling step, little literature delves deeply into the performance dominance and limitations of the two models. In this paper, we first theoretically and empirically demonstrate that the local model achieves higher re-labeling accuracy over local data while the global model can progressively improve the re-labeling performance by introducing the extra data knowledge of other clients. Based on these findings, we propose BSemiFL which re-labels the local data through the collaboration between the local and global model in a Bayesian approach. Specifically, to re-label any given local sample, BSemiFL first uses Bayesian inference to assess the closeness of the local/global model to the sample. Then, it applies a weighted combination of their pseudo labels, using the closeness as the weights. Theoretical analysis shows that the labeling error of our method is smaller than that of simply using the global model, the local model, or their simple average. Experimental results show that BSemiFL improves the performance by up to $9. 8%$ as compared to state-of-the-art methods.

Details

NeurIPS Conference 2025 Conference Paper

GUIDED: Granular Understanding via Identification, Detection, and Discrimination for Fine-Grained Open-Vocabulary Object Detection

Jiaming Li
Zhijia Liang
Weikai Chen
Lin Ma
Guanbin Li

Fine-grained open-vocabulary object detection (FG-OVD) aims to detect novel object categories described by attribute-rich texts. While existing open-vocabulary detectors show promise at the base-category level, they underperform in fine-grained settings due to the semantic entanglement of subjects and attributes in pretrained vision-language model (VLM) embeddings -- leading to over-representation of attributes, mislocalization, and semantic drift in embedding space. We propose GUIDED, a decomposition framework specifically designed to address the semantic entanglement between subjects and attributes in fine-grained prompts. By separating object localization and fine-grained recognition into distinct pathways, GUIDED aligns each subtask with the module best suited for its respective roles. Specifically, given a fine-grained class name, we first use a language model to extract a coarse-grained subject and its descriptive attributes. Then the detector is guided solely by the subject embedding, ensuring stable localization unaffected by irrelevant or overrepresented attributes. To selectively retain helpful attributes, we introduce an attribute embedding fusion module that incorporates attribute information into detection queries in an attention-based manner. This mitigates over-representation while preserving discriminative power. Finally, a region-level attribute discrimination module compares each detected region against full fine-grained class names using a refined vision-language model with a projection head for improved alignment. Extensive experiments on FG-OVD and 3F-OVD benchmarks show that GUIDED achieves new state-of-the-art results, demonstrating the benefits of disentangled modeling and modular optimization.

PDF Details

AAAI Conference 2025 Conference Paper

Hierarchical Context Pruning: Optimizing Real-World Code Completion with Repository-Level Pretrained Code LLMs

Lei Zhang
Yunshui Li
Jiaming Li
Xiaobo Xia
Jiaxi Yang
Run Luo
Minzheng Wang
Longze Chen

Some of the latest released Code Large Language Models (Code LLMs) have been trained on repository-level code data, enabling them to perceive repository structures and utilize cross-file code information. This capability allows us to directly concatenate the content of repository code files in prompts to achieve repository-level code completion. However, in real development scenarios, directly concatenating all code repository files in a prompt can easily exceed the context window of Code LLMs, leading to a significant decline in completion performance. Additionally, overly long prompts can increase completion latency, negatively impacting the user experience. In this study, we conducted extensive experiments, including completion error analysis, topology dependency analysis, and cross-file content analysis, to investigate the factors affecting repository-level code completion. Based on the conclusions drawn from these preliminary experiments, we proposed a strategy called **Hierarchical Context Pruning (HCP)** to construct high-quality completion prompts. We applied the **HCP** to six Code LLMs and evaluated them on the CrossCodeEval dataset. The experimental results showed that, compared to previous methods, the prompts constructed using our **HCP** strategy achieved higher completion accuracy on five out of six Code LLMs. Additionally, the **HCP** managed to keep the prompt length around 8k tokens (whereas the full repository code is approximately 50k tokens), significantly improving completion throughput. Our code and data will be publicly available.

PDF Details DOI

AAAI Conference 2025 Conference Paper

IDseq: Decoupled and Sequentially Detecting and Grounding Multi-Modal Media Manipulation

Runxin Liu
Tian Xie
Jiaming Li
Lingyun Yu
Hongtao Xie

Detecting and grounding multi-modal media manipulation aims to categorize the type and localize the region of manipulation for image-text pairs in both two modalities. Existing methods have not sufficiently explored the intrinsic properties of the manipulated images, which contain both forgery and content features, leading to inefficient utilization. To address this problem, we propose an Image-Driven Decoupled Sequential Framework (IDseq), designed to decouple image features and rationally integrate them to accomplish different sub-tasks effectively. Specifically, IDseq employs two specially designed disentangled losses to guide the disentangled learning of forgery and content features. To efficiently leverage these features, we propose a Decoupled Image Manipulation Decoder (DIMD) that processes image tasks within a decoupled schema. We mitigate their exclusive competition by separating the image tasks into forgery-relevant and content-relevant components and training them without gradient interaction. Additionally, we utilize content features enhanced by the proposed Manipulation Indicator Generator (MIG) for the text tasks, which provide the maximal visual information as a reference while eliminating interference from unverified image data. Extensive experiments show the superiority of our IDseq, where it notably outperforms SOTA methods on the fine-grained classification by 3.8% in mAP and the forgery face grounding by 8.7% in IoUmean, even 1.3% in F1 on the most challenging manipulated text grounding.

PDF Details DOI

EAAI Journal 2025 Journal Article

Improving model calibration in bone marrow cell classification through mixup and center loss fusion

Shuming Cheng
Qinghang Lu
Qianhang Guo
Yunqi Lin
Mingxin Li
Xingyu Zhao
Liang Guo
Jiaming Li

The analysis of bone marrow cell morphology is essential for the accurate diagnosis of hematological disorders. Traditional manual classification methods are time-consuming and labor-intensive. Although current automatic deep learning techniques mitigate these issues, they may still present significant risks in critical medical diagnostics due to overconfidence in predictions. To tackle these challenges, this paper proposes a novel calibration method called MixCL (Mix-Center Loss). MixCL combines the simple and effective data augmentation method Mixup with deep metric learning Center Loss, achieved through the design of a new loss function. By utilizing Mixup to generate mixing centers that enrich the feature sampling in the feature space, and leveraging the clustering effect of Center Loss to enhance the grouping of similar samples, MixCL combines the strengths of both methods. The effectiveness of MixCL is validated using three real bone marrow cell image datasets, demonstrating significant reductions in Expected Calibration Error (ECE) and Overconfidence Error (OE) for in-distribution samples. For example, in Shifted Windows Transformer model, ECE and OE metrics decreased across all datasets, with reductions averaging 1. 72% in ECE and 2. 10% in OE. The confidence Kernel Density Estimation (KDE) plot reveals that models using MixCL more effectively manage uncertainty in out-of-distribution samples, ensuring better differentiation between in-distribution and out-of-distribution samples. Thus, the proposed method effectively improves the calibration performance of the model while exhibiting better generalization performance, significantly improved when compared with current advanced bone marrow cell classification methods. Moreover, it has potential applications in various image classification fields, providing reliable confidence estimates.

Details DOI

NeurIPS Conference 2025 Conference Paper

OpenOmni: Advancing Open-Source Omnimodal Large Language Models with Progressive Multimodal Alignment and Real-time Emotional Speech Synthesis

Run Luo
Ting-En Lin
Haonan Zhang
Yuchuan Wu
Xiong Liu
Yongbin Li
Longze Chen
Jiaming Li

Recent advancements in omnimodal learning have significantly improved understanding and generation across images, text, and speech, yet these developments remain predominantly confined to proprietary models. The lack of high-quality omnimodal datasets and the challenges of real-time emotional speech synthesis have notably hindered progress in open-source research. To address these limitations, we introduce OpenOmni, a two-stage training framework that integrates omnimodal alignment and speech generation to develop a state-of-the-art omnimodal large language model. In the alignment phase, a pretrained speech model undergoes further training on image-text tasks, enabling (near) zero-shot generalization from vision to speech, outperforming models trained on tri-modal datasets. In the speech generation phase, a lightweight decoder is trained on speech tasks with direct preference optimization, which enables real-time emotional speech synthesis with high fidelity. Extensive experiments demonstrate that OpenOmni surpasses state-of-the-art models across omnimodal, vision-language, and speech-language benchmarks. It achieves a 4-point absolute improvement on OmniBench over the leading open-source model VITA, despite using 5$\times$ fewer training examples and a smaller model size (7B vs. 7$\times$8B). Besides, OpenOmni achieves real-time speech generation with less than 1 second latency at non-autoregressive mode, reducing inference time by 5$\times$ compared to autoregressive methods, and improves emotion classification accuracy by 7. 7\%. The codebase is available at https: //github. com/RainBowLuoCS/OpenOmni.

PDF Details

EAAI Journal 2024 Journal Article

A novel bone marrow cell recognition method based on multi-scale information and reject option

Haisen He
Zilan Li
Yunqi Lin
Tongyi Wei
Qianghang Guo
Qinghang Lu
Liang Guo
Qingmao Zhang

The morphology of bone marrow cells is crucial for the diagnostics of blood disorders. However, the traditional approach of cell sorting and counting under a microscope is time-consuming and error-prone. To assist experts in diagnosis, there is considerable interest at present in using neural networks to develop automatic recognition algorithms. However, the results obtained from these approaches indicate that the achieved accuracy is not entirely convincing, and there remains a risk of misdiagnosis. In fact, a reliable collaborative recognition algorithm not only needs to have high recognition accuracy but also must be able to confidently reject low-confidence classification results and leave them for human experts to handle. Therefore, this paper proposes a method that combines high accuracy with a rejection recognition option for bone marrow cell recognition. This method incorporates Crossformer (Cross) and Class Centroid Learning (CCL), which integrate multi-scale information from cell images, enhance image feature differentiation, and effectively reject low-confidence predictions. In the experiment, the recognition accuracy of Cross-CCL reaches 94. 41%, surpassing the performance of the current mainstream and advanced methods for bone marrow cell recognition. Simultaneously, CCL serves as a dependable collaborative algorithm by effectively rejecting 21. 07% of cell images that represent low-confidence prediction results. Moreover, the accuracy of identification for the remaining portions exceeds 98%. This rejection mechanism not only enhances the efficiency of expert resource allocation but also ensures a higher level of confidence in the identified cell images.

Details DOI

NeurIPS Conference 2024 Conference Paper

II-Bench: An Image Implication Understanding Benchmark for Multimodal Large Language Models

Ziqiang Liu
Feiteng Fang
Xi Feng
Xinrun Du
Chenhao Zhang
Zekun Wang
yuelin bai
Qixuan Zhao

The rapid advancements in the development of multimodal large language models (MLLMs) have consistently led to new breakthroughs on various benchmarks. In response, numerous challenging and comprehensive benchmarks have been proposed to more accurately assess the capabilities of MLLMs. However, there is a dearth of exploration of the higher-order perceptual capabilities of MLLMs. To fill this gap, we propose the Image Implication understanding Benchmark, II-Bench, which aims to evaluate the model's higher-order perception of images. Through extensive experiments on II-Bench across multiple MLLMs, we have made significant findings. Initially, a substantial gap is observed between the performance of MLLMs and humans on II-Bench. The pinnacle accuracy of MLLMs attains 74. 8%, whereas human accuracy averages 90%, peaking at an impressive 98%. Subsequently, MLLMs perform worse on abstract and complex images, suggesting limitations in their ability to understand high-level semantics and capture image details. Finally, it is observed that most models exhibit enhanced accuracy when image sentiment polarity hints are incorporated into the prompts. This observation underscores a notable deficiency in their inherent understanding of image sentiment. We believe that II-Bench will inspire the community to develop the next generation of MLLMs, advancing the journey towards expert artificial general intelligence (AGI). II-Bench is publicly available at https: //huggingface. co/datasets/m-a-p/II-Bench.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Instruction-Guided Visual Masking

Jinliang Zheng
Jianxiong Li
Sijie Cheng
Yinan Zheng
Jiaming Li
Jihao Liu
Yu Liu
Jingjing Liu

Instruction following is crucial in contemporary LLM. However, when extended to multimodal setting, it often suffers from misalignment between specific textual instruction and targeted local region of an image. To achieve more accurate and nuanced multimodal instruction following, we introduce Instruction-guided Visual Masking (IVM), a new versatile visual grounding model that is compatible with diverse multimodal models, such as LMM and robot model. By constructing visual masks for instruction-irrelevant regions, IVM-enhanced multimodal models can effectively focus on task-relevant image regions to better align with complex instructions. Specifically, we design a visual masking data generation pipeline and create an IVM-Mix-1M dataset with 1 million image-instruction pairs. We further introduce a new learning technique, Discriminator Weighted Supervised Learning (DWSL) for preferential IVM training that prioritizes high-quality data samples. Experimental results on generic multimodal tasks such as VQA and embodied robotic control demonstrate the versatility of IVM, which as a plug-and-play tool, significantly boosts the performance of diverse multimodal models, yielding new state-of-the-art results across challenging multimodal benchmarks. Code, model and data are available at https: //github. com/2toinf/IVM.

PDF Details DOI

YNICL Journal 2024 Journal Article

Large-scale effective connectivity analysis reveals the existence of two mutual inhibitory systems in patients with major depression

Jia Wang
Baojuan Li
Jian Liu
Jiaming Li
Adeel Razi
Kaizhong Zheng
Baoyu Yan
Huaning Wang

It is posited that cognitive and affective dysfunction in patients with major depression disorder (MDD) may be caused by dysfunctional signal propagation in the brain. By leveraging dynamic causal modeling, we investigated large-scale directed signal propagation (effective connectivity) among distributed large-scale brain networks with 43 MDD patients and 56 healthy controls. The results revealed the existence of two mutual inhibitory systems: the anterior default mode network, auditory network, sensorimotor network, salience network and visual networks formed an "emotional" brain, while the posterior default mode network, central executive networks, cerebellum and dorsal attention network formed a "rational brain". These two networks exhibited excitatory intra-system connectivity and inhibitory inter-system connectivity. Patients were characterized by potentiated intra-system connections within the "emotional/sensory brain", as well as over-inhibition of the "rational brain" by the "emotional/sensory brain". The hierarchical architecture of the large-scale effective connectivity networks was then analyzed using a PageRank algorithm which revealed a shift of the controlling role of the "rational brain" to the "emotional/sensory brain" in the patients. These findings inform basic organization of distributed large-scale brain networks and furnish a better characterization of the neural mechanisms of depression, which may facilitate effective treatment.

Details DOI