Visual instruction tuning is a key training stage of large multimodal models. However, when learning multiple visual tasks simultaneously, this approach often results in suboptimal and imbalanced overall performance due to latent knowledge conflicts across tasks. To mitigate this issue, we propose a novel Adaptive Task Balancing approach tailored for visual instruction tuning (VisATB). Specifically, we measure two critical dimensions for visual task balancing based on validation performance: (1) Inter-Task Contribution, the mechanism where learning one task enhances the performance on others owing to shared knowledge across tasks, and (2) Intra-Task Difficulty, which denotes the inherent learning difficulty of a single task. Furthermore, we propose prioritizing three categories of tasks with greater weight: those that offer substantial contributions to others, those that receive minimal contributions from others, and those that present high learning difficulties. Among these three task weighting strategies, the first and third focus on improving overall performance, and the second targets the mitigation of performance imbalance. Extensive experiments on three benchmarks demonstrate that our VisATB approach consistently achieves superior and more balanced overall performance in visual instruction tuning. The data, code, and models are available at https://github.com/YanqiDai/VisATB.

Composed Image Retrieval (CIR) aims to retrieve target images from candidate set using a hybrid-modality query consisting of a reference image and a relative caption that describes the user intent. Recent studies attempt to utilize Vision-Language Pre-training Models (VLPMs) with various fusion strategies for addressing the task. However, these methods typically fail to simultaneously meet two key requirements of CIR: comprehensively extracting visual information and faithfully following the user intent. In this work, we propose CIR-LVLM, a novel framework that leverages the large vision-language model (LVLM) as the powerful user intent-aware encoder to better meet these requirements. Our motivation is to explore the advanced reasoning and instruction-following capabilities of LVLM for accurately understanding and responding the user intent. Furthermore, we design a novel hybrid intent instruction module to provide explicit intent guidance at two levels: (1) The task prompt clarifies the task requirement and assists the model in discerning user intent at the task level. (2) The instance-specific soft prompt, which is adaptively selected from the learnable prompt pool, enables the model to better comprehend the user intent at the instance level compared to a universal prompt for all instances. CIR-LVLM achieves state-of-the-art performance across three prominent benchmarks with acceptable inference efficiency. We believe this study provides fundamental insights into CIR-related fields.

Contrastive Language-Image Pre-training (CLIP) achieves impressive performance on tasks like image classification and image-text retrieval by learning on large-scale image-text datasets. However, CLIP struggles with dense prediction tasks due to the poor grasp of the fine-grained details. Although existing works pay attention to this issue, they achieve limited improvements and usually sacrifice the important visual-semantic consistency. To overcome these limitations, we propose FineCLIP, which keeps the global contrastive learning to preserve the visual-semantic consistency and further enhances the fine-grained understanding through two innovations: 1) A real-time self-distillation scheme that facilitates the transfer of representation capability from global to local features. 2) A semantically-rich regional contrastive learning paradigm with generated region-text pairs, boosting the local representation capabilities with abundant fine-grained knowledge. Both cooperate to fully leverage diverse semantics and multi-grained complementary information. To validate the superiority of our FineCLIP and the rationality of each design, we conduct extensive experiments on challenging dense prediction and image-level tasks. All the observations demonstrate the effectiveness of FineCLIP.

Data sharing in the field of industrial Internet provides a powerful impetus for the vigorous development of industrial Internet by integrating management and flexible configuration of resources. However, there are single point of failure and data leakage problems in the current industrial Internet data sharing and storage process due to the adoption of single-cloud storage mode. This paper proposes a multi-cloud collaborative data security sharing scheme with blockchain indexing. The blockchain-assisted multi-cloud storage model is constructed to enable indirect retrieval of industrial data through a federation chain. Enterprises on the alliance chain cannot directly access the industrial data stored in their respective private clouds for each other. To address the lack of trustworthiness of ciphertext retrieval, a tree-structured indexing method supporting keyword ciphertext retrieval is designed to realize efficient keyword token search. To prevent unfair trading behavior of data sharing participants, a fair trading contract is designed to overcome the disadvantage of dishonest execution of protocols by cloud servers. The security analysis shows that the scheme satisfies IND-CKA security. Simulation results show that the tree indexing approach constructed in this paper has less time overhead compared to the traditional sequential indexing and other robust technique. And the efficiency advantage is more prominent when the number of keywords is huge.

Existing light field based works utilize either views or focal stacks for saliency detection. However, since depth information exists implicitly in adjacent views or different focal slices, it is difficult to exploit scene depth information from both. By comparison, Epipolar Plane Images (EPIs) provide explicit accurate scene depth and occlusion information by projected pixel lines. Due to the fact that the depth of an object is often continuous, the distribution of occlusion edges concentrates more on object boundaries compared with traditional color edges, which is more beneficial for improving accuracy and completeness of saliency detection. In this paper, we propose a learning-based network to exploit occlusion features from EPIs and integrate high-level features from the central view for accurate salient object detection. Specifically, a novel Occlusion Extraction Module is proposed to extract occlusion boundary features from horizontal and vertical EPIs. In order to naturally combine occlusion features in EPIs and high-level features in central view, we design a concise Bi-directional Guiding Flow based on cascaded decoders. The flow leverages generated salient edge predictions and salient object predictions to refine features in mutual encoding processes. Experimental results demonstrate that our approach achieves state-of-the-art performance in both segmentation accuracy and edge clarity.