Yi Ding

Yi Ding, Lijun Li, Bing Cao et al.

Large Vision-Language Models (VLMs) have achieved remarkable performance across a wide range of tasks. However, their deployment in safety-critical domains poses significant challenges. Existing safety fine-tuning methods, which focus on textual or multimodal content, fall short in addressing challenging cases or disrupt the balance between helpfulness and harmlessness. Our evaluation highlights a safety reasoning gap: these methods lack safety visual reasoning ability, leading to such bottlenecks. To address this limitation and enhance both visual perception and reasoning in safety-critical contexts, we propose a novel dataset that integrates multi-image inputs with safety Chain-of-Thought (CoT) labels as fine-grained reasoning logic to improve model performance. Specifically, we introduce the Multi-Image Safety (MIS) dataset, an instruction-following dataset tailored for multi-image safety scenarios, consisting of training and test splits. Our experiments demonstrate that fine-tuning InternVL2.5-8B with MIS significantly outperforms both powerful open-source models and API-based models in challenging multi-image tasks requiring safety-related visual reasoning. This approach not only delivers exceptional safety performance but also preserves general capabilities without any trade-offs. Specifically, fine-tuning with MIS increases average accuracy by 0.83% across five general benchmarks and reduces the Attack Success Rate (ASR) on multiple safety benchmarks by a large margin.

Ziqi Miao, Yi Ding, Lijun Li et al.

With the emergence of strong vision language capabilities, multimodal large language models (MLLMs) have demonstrated tremendous potential for real-world applications. However, the security vulnerabilities exhibited by the visual modality pose significant challenges to deploying such models in open-world environments. Recent studies have successfully induced harmful responses from target MLLMs by encoding harmful textual semantics directly into visual inputs. However, in these approaches, the visual modality primarily serves as a trigger for unsafe behavior, often exhibiting semantic ambiguity and lacking grounding in realistic scenarios. In this work, we define a novel setting: vision-centric jailbreak, where visual information serves as a necessary component in constructing a complete and realistic jailbreak context. Building on this setting, we propose the VisCo (Visual Contextual) Attack. VisCo fabricates contextual dialogue using four distinct vision-focused strategies, dynamically generating auxiliary images when necessary to construct a vision-centric jailbreak scenario. To maximize attack effectiveness, it incorporates automatic toxicity obfuscation and semantic refinement to produce a final attack prompt that reliably triggers harmful responses from the target black-box MLLMs. Specifically, VisCo achieves a toxicity score of 4.78 and an Attack Success Rate (ASR) of 85% on MM-SafetyBench against GPT-4o, significantly outperforming the baseline, which achieves a toxicity score of 2.48 and an ASR of 22.2%. Code: https://github.com/Dtc7w3PQ/Visco-Attack.

Bing Cao, Yinan Xia, Yi Ding et al.

The inherent challenge of image fusion lies in capturing the correlation of multi-source images and comprehensively integrating effective information from different sources. Most existing techniques fail to perform dynamic image fusion while notably lacking theoretical guarantees, leading to potential deployment risks in this field. Is it possible to conduct dynamic image fusion with a clear theoretical justification? In this paper, we give our solution from a generalization perspective. We proceed to reveal the generalized form of image fusion and derive a new test-time dynamic image fusion paradigm. It provably reduces the upper bound of generalization error. Specifically, we decompose the fused image into multiple components corresponding to its source data. The decomposed components represent the effective information from the source data, thus the gap between them reflects the Relative Dominability (RD) of the uni-source data in constructing the fusion image. Theoretically, we prove that the key to reducing generalization error hinges on the negative correlation between the RD-based fusion weight and the uni-source reconstruction loss. Intuitively, RD dynamically highlights the dominant regions of each source and can be naturally converted to the corresponding fusion weight, achieving robust results. Extensive experiments and discussions with in-depth analysis on multiple benchmarks confirm our findings and superiority. Our code is available at https://github.com/Yinan-Xia/TTD.

Jintu Zheng, Yi Ding, Qizhe Liu et al.

Traditional fluorescence staining is phototoxic to live cells, slow, and expensive; thus, the subcellular structure prediction (SSP) from transmitted light (TL) images is emerging as a label-free, faster, low-cost alternative. However, existing approaches utilize 3D networks for one-to-one voxel level dense prediction, which necessitates a frequent and time-consuming Z-axis imaging process. Moreover, 3D convolutions inevitably lead to significant computation and GPU memory overhead. Therefore, we propose an efficient framework, SparseSSP, predicting fluorescent intensities within the target voxel grid in an efficient paradigm instead of relying entirely on 3D topologies. In particular, SparseSSP makes two pivotal improvements to prior works. First, SparseSSP introduces a one-to-many voxel mapping paradigm, which permits the sparse TL slices to reconstruct the subcellular structure. Secondly, we propose a hybrid dimensions topology, which folds the Z-axis information into channel features, enabling the 2D network layers to tackle SSP under low computational cost. We conduct extensive experiments to validate the effectiveness and advantages of SparseSSP on diverse sparse imaging ratios, and our approach achieves a leading performance compared to pure 3D topologies. SparseSSP reduces imaging frequencies compared to previous dense-view SSP (i.e., the number of imaging is reduced up to 87.5% at most), which is significant in visualizing rapid biological dynamics on low-cost devices and samples.

Xinliang Zhou, Chenyu Liu, Zhisheng Chen et al.

Brain foundation models (BFMs) have emerged as a transformative paradigm in computational neuroscience, offering a revolutionary framework for processing diverse neural signals across different brain-related tasks. These models leverage large-scale pre-training techniques, allowing them to generalize effectively across multiple scenarios, tasks, and modalities, thus overcoming the traditional limitations faced by conventional artificial intelligence (AI) approaches in understanding complex brain data. By tapping into the power of pretrained models, BFMs provide a means to process neural data in a more unified manner, enabling advanced analysis and discovery in the field of neuroscience. In this survey, we define BFMs for the first time, providing a clear and concise framework for constructing and utilizing these models in various applications. We also examine the key principles and methodologies for developing these models, shedding light on how they transform the landscape of neural signal processing. This survey presents a comprehensive review of the latest advancements in BFMs, covering the most recent methodological innovations, novel views of application areas, and challenges in the field. Notably, we highlight the future directions and key challenges that need to be addressed to fully realize the potential of BFMs. These challenges include improving the quality of brain data, optimizing model architecture for better generalization, increasing training efficiency, and enhancing the interpretability and robustness of BFMs in real-world applications.

Yi Ding, Qiqi Yang, Guozheng Wu et al.

The identification of nerve is difficult as structures of nerves are challenging to image and to detect in ultrasound images. Nevertheless, the nerve identification in ultrasound images is a crucial step to improve performance of regional anesthesia. In this paper, a network called Brachial Plexus Multi-instance Segmentation Network (BPMSegNet) is proposed to identify different tissues (nerves, arteries, veins, muscles) in ultrasound images. The BPMSegNet has three novel modules. The first is the spatial local contrast feature, which computes contrast features at different scales. The second one is the self-attention gate, which reweighs the channels in feature maps by their importance. The third is the addition of a skip concatenation with transposed convolution within a feature pyramid network. The proposed BPMSegNet is evaluated by conducting experiments on our constructed Ultrasound Brachial Plexus Dataset (UBPD). Quantitative experimental results show the proposed network can segment multiple tissues from the ultrasound images with a good performance.

Yi Ding, Wei Zheng, Guozheng Wu et al.

When diagnosing the brain tumor, doctors usually make a diagnosis by observing multimodal brain images from the axial view, the coronal view and the sagittal view, respectively. And then they make a comprehensive decision to confirm the brain tumor based on the information obtained from multi-views. Inspired by this diagnosing process and in order to further utilize the 3D information hidden in the dataset, this paper proposes a multi-view dynamic fusion framework to improve the performance of brain tumor segmentation. The proposed framework consists of 1) a multi-view deep neural network architecture, which represents multi learning networks for segmenting the brain tumor from different views and each deep neural network corresponds to multi-modal brain images from one single view and 2) the dynamic decision fusion method, which is mainly used to fuse segmentation results from multi-views as an integrate one and two different fusion methods, the voting method and the weighted averaging method, have been adopted to evaluate the fusing process. Moreover, the multi-view fusion loss, which consists of the segmentation loss, the transition loss and the decision loss, is proposed to facilitate the training process of multi-view learning networks so as to keep the consistency of appearance and space, not only in the process of fusing segmentation results, but also in the process of training the learning network. \par By evaluating the proposed framework on BRATS 2015 and BRATS 2018, it can be found that the fusion results from multi-views achieve a better performance than the segmentation result from the single view and the effectiveness of proposed multi-view fusion loss has also been proved. Moreover, the proposed framework achieves a better segmentation performance and a higher efficiency compared to other counterpart methods.

Yi Ding, Fuyuan Tan, Zhen Qin et al.

The need for medical image encryption is increasingly pronounced, for example to safeguard the privacy of the patients' medical imaging data. In this paper, a novel deep learning-based key generation network (DeepKeyGen) is proposed as a stream cipher generator to generate the private key, which can then be used for encrypting and decrypting of medical images. In DeepKeyGen, the generative adversarial network (GAN) is adopted as the learning network to generate the private key. Furthermore, the transformation domain (that represents the "style" of the private key to be generated) is designed to guide the learning network to realize the private key generation process. The goal of DeepKeyGen is to learn the mapping relationship of how to transfer the initial image to the private key. We evaluate DeepKeyGen using three datasets, namely: the Montgomery County chest X-ray dataset, the Ultrasonic Brachial Plexus dataset, and the BraTS18 dataset. The evaluation findings and security analysis show that the proposed key generation network can achieve a high-level security in generating the private key.

Tian Lan, Yuanyuan Li, Jonah Kimani Murugi et al.

The early detection and early diagnosis of lung cancer are crucial to improve the survival rate of lung cancer patients. Pulmonary nodules detection results have a significant impact on the later diagnosis. In this work, we propose a new network named RUN to complete nodule detection in a single step by bypassing the candidate selection. The system introduces the shortcut of the residual network to improve the traditional U-Net, thereby solving the disadvantage of poor results due to its lack of depth. Furthermore, we compare the experimental results with the traditional U-Net. We validate our method in LUng Nodule Analysis 2016 (LUNA16) Nodule Detection Challenge. We acquire a sensitivity of 90.90% at 2 false positives per scan and therefore achieve better performance than the current state-of-the-art approaches.