❌

Normal view

Received — 27 May 2026 ⏭

DisDop: Distillation with Domain Priors for Open-Vocabulary Aerial Object Detection

✇cs.AI, q-bio.NC updates on arXiv.org
By: Ruihao Xu · Yong Liu · Yansong Tang · Sule Bai · Xubing Ye · Bingyao Yu · Yutao Guo · Jiwen Lu · Jie Zhou
26 May 2026 at 12:00
arXiv:2605.24639v1 Announce Type: cross Abstract: With the widespread application of drones in recent years, object detection of aerial images has attracted increasing attention, especially open-vocabulary aerial detection which is not restricted to predefined categories. Due to the scarcity of drone's viewpoint images and their significant differences from natural images, it is difficult to achieve satisfying results by directly applying vanilla open-vocabulary detection methods designed for natural scenarios. Some studies propose to transfer knowledge from pre-trained models by using lightweight networks or generating pseudo labels, but they tend to rely on models trained on natural images, neglecting the potential of foundation models specifically tailored for remote sensing and aerial imagery. To address this limitation, we propose DisDop, a unified framework that systematically distills multi-level domain priors from remote sensing foundation models (e.g., RemoteCLIP and DINOv3) into a lightweight detector. Specifically, we first distill visual priors through a teacher fusion strategy that combines RemoteCLIP's cross-modal alignment capability with DINOv3's fine-grained local feature extraction ability, transferring their complementary strengths to the detector's backbone. Second, we distill textual priors embedded in RemoteCLIP's text encoder by explicitly modeling inter-category semantic relationships, while incorporating global contextual priors to enhance local feature representation for small objects. Through this multi-level prior distillation framework, our DisDop achieves new state-of-the-art performance on open-vocabulary aerial detection benchmarks. Extensive ablation analysis also demonstrates the rationality and effectiveness of our proposed modules.

Safety in Embodied AI: A Survey of Risks, Attacks, and Defenses

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xiao Li · Xiang Zheng · Yifeng Gao · Xinyu Xia · Yixu Wang · Xin Wang · Ye Sun · Yunhan Zhao · Ming Wen · Jiayu Li · Zixing Chen · Xun Gong · Yi Liu · Yige Li · Yutao Wu · Cong Wang · Jun Sun · Yixin Cao · Zhineng Chen · Jingjing Chen · Tao Gui · Qi Zhang · Zuxuan Wu · Xipeng Qiu · Xuanjing Huang · Tiehua Zhang · Zhipeng Wei · Kun Wang · Xinfeng Li · Hanxun Huang · Sarah Erfani · James Bailey · Jianping Wang · Chaowei Xiao · Ran He · Bo Li · Xingjun Ma · Yu-Gang Jiang
26 May 2026 at 12:00
arXiv:2605.02900v2 Announce Type: replace-cross Abstract: Embodied Artificial Intelligence (Embodied AI) integrates perception, cognition, planning, and interaction into agents that operate in open-world, safety-critical environments. As these systems gain autonomy and enter domains such as transportation, healthcare, and industrial or assistive robotics, ensuring their safety becomes both technically challenging and socially indispensable. Unlike digital AI systems, embodied agents must act under uncertain sensing, incomplete knowledge, and dynamic human-robot interactions, where failures can directly lead to physical harm. This survey provides a comprehensive and structured review of safety research in embodied AI, examining attacks and defenses across the full embodied pipeline, from perception and cognition to planning, action and interaction, and agentic system. We introduce a multi-level taxonomy that unifies fragmented lines of work and connects embodied-specific safety findings with broader advances in vision, language, and multimodal foundation models. Our review synthesizes insights from over 500 papers spanning adversarial, backdoor, jailbreak, and hardware-level attacks; attack detection, safe training and robust inference; and risk-aware human-agent interaction. This analysis reveals several overlooked challenges, including the fragility of multimodal perception fusion, the instability of planning under jailbreak attacks, and the trustworthiness of human-agent interaction in open-ended scenarios. By organizing the field into a coherent framework and identifying critical research gaps, this survey provides a roadmap for building embodied agents that are not only capable and autonomous but also safe, robust, and reliable in real-world deployment.
Received — 2 April 2026 ⏭

Dunhuang Daxiefei Decoction ameliorates acute lung injury via the HIF-1alpha/glycolysis/H3K18la axis

✇Omics In Lung
By: Jiayun Wang · Liangbo Jiao · Zhiqiang Yan · Shun Wang · Suwan Wu · Tao Guo · Yuetong Yang · Xiaoyingzi Su · Bo Tan · Aidong Yang
28 March 2026 at 18:00

J Ethnopharmacol. 2026 Mar 26;365:121591. doi: 10.1016/j.jep.2026.121591. Online ahead of print.

ABSTRACT

ETHNOPHARMACOLOGICAL RELEVANCE: Acute lung injury (ALI) lacks effective therapies. HIF-1α-driven glycolysis can promote histone lactylation and sustain pro-inflammatory (M1) macrophage responses. Daxiefei Decoction (DXFD), a classic traditional Chinese medicine formula, is used for pulmonary inflammatory diseases, but its immunometabolic mechanism remains unclear.

AIM OF THE STUDY: To evaluate the protective efficacy of DXFD against lipopolysaccharide (LPS)-induced ALI and to determine whether it acts through the HIF-1α/glycolysis/histone H3K18 lactylation (H3K18la) axis to regulate macrophage polarization.

MATERIALS & METHODS: DXFD constituents were characterized by UPLC-LTQ-Orbitrap-MS/MS, followed by network pharmacology, molecular docking, and molecular dynamics (MD) simulations. Lung transcriptomics and metabolomics were performed in ALI mice. Efficacy and mechanisms were assessed in LPS-challenged mice and RAW264.7 macrophages using histopathology, ELISA, qRT-PCR, Western blotting, and immunofluorescence. HIF-1α overexpression was used for validation.

RESULTS: DXFD dose-dependently alleviated lung injury and reduced pro-inflammatory cytokines in vivo, and suppressed M1 polarization in vivo and in LPS-stimulated macrophages. Multi-omics indicated activation of HIF-1α-associated inflammatory and glycolytic programs in ALI, which were normalized by DXFD. DXFD decreased glycolytic enzyme expression and reduced histone H3K18 lactylation (H3K18la); these effects were partially reversed by HIF-1α overexpression. Molecular docking and dynamics suggested stable binding of baicalin to HIF-1α.

CONCLUSIONS: DXFD mitigates ALI by dampening HIF-1α-dependent glycolysis and H3K18la, thereby restraining M1-driven inflammatory amplification.

PMID:41903585 | DOI:10.1016/j.jep.2026.121591

Received — 26 March 2026 ⏭

JFTA-Bench: Evaluate LLM's Ability of Tracking and Analyzing Malfunctions Using Fault Trees

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yuhui Wang · Zhixiong Yang · Ming Zhang · Shihan Dou · Zhiheng Xi · Enyu Zhou · Senjie Jin · Yujiong Shen · Dingwei Zhu · Yi Dong · Tao Gui · Qi Zhang · Xuanjing Huang
25 March 2026 at 12:00
arXiv:2603.22978v1 Announce Type: new Abstract: In the maintenance of complex systems, fault trees are used to locate problems and provide targeted solutions. To enable fault trees stored as images to be directly processed by large language models, which can assist in tracking and analyzing malfunctions, we propose a novel textual representation of fault trees. Building on it, we construct a benchmark for multi-turn dialogue systems that emphasizes robust interaction in complex environments, evaluating a model's ability to assist in malfunction localization, which contains $3130$ entries and $40.75$ turns per entry on average. We train an end-to-end model to generate vague information to reflect user behavior and introduce long-range rollback and recovery procedures to simulate user error scenarios, enabling assessment of a model's integrated capabilities in task tracking and error recovery, and Gemini 2.5 pro archives the best performance.
Received — 5 March 2026 ⏭

Kaleido: Open-Sourced Multi-Subject Reference Video Generation Model

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhenxing Zhang · Jiayan Teng · Zhuoyi Yang · Tiankun Cao · Cheng Wang · Xiaotao Gu · Jie Tang · Dan Guo · Meng Wang
5 March 2026 at 13:00
arXiv:2510.18573v2 Announce Type: replace-cross Abstract: We present Kaleido, a subject-to-video~(S2V) generation framework, which aims to synthesize subject-consistent videos conditioned on multiple reference images of target subjects. Despite recent progress in S2V generation models, existing approaches remain inadequate at maintaining multi-subject consistency and at handling background disentanglement, often resulting in lower reference fidelity and semantic drift under multi-image conditioning. These shortcomings can be attributed to several factors. Primarily, the training dataset suffers from a lack of diversity and high-quality samples, as well as cross-paired data, i.e., paired samples whose components originate from different instances. In addition, the current mechanism for integrating multiple reference images is suboptimal, potentially resulting in the confusion of multiple subjects. To overcome these limitations, we propose a dedicated data construction pipeline, incorporating low-quality sample filtering and diverse data synthesis, to produce consistency-preserving training data. Moreover, we introduce Reference Rotary Positional Encoding (R-RoPE) to process reference images, enabling stable and precise multi-image integration. Extensive experiments across numerous benchmarks demonstrate that Kaleido significantly outperforms previous methods in consistency, fidelity, and generalization, marking an advance in S2V generation.

Social-JEPA: Emergent Geometric Isomorphism

✇cs.AI, q-bio.NC updates on arXiv.org
By: Haoran Zhang · Youjin Wang · Yi Duan · Rong Fu · Dianyu Zhao · Sicheng Fan · Shuaishuai Cao · Wentao Guo · Xiao Zhou
4 March 2026 at 13:00
arXiv:2603.02263v1 Announce Type: cross Abstract: World models compress rich sensory streams into compact latent codes that anticipate future observations. We let separate agents acquire such models from distinct viewpoints of the same environment without any parameter sharing or coordination. After training, their internal representations exhibit a striking emergent property: the two latent spaces are related by an approximate linear isometry, enabling transparent translation between them. This geometric consensus survives large viewpoint shifts and scant overlap in raw pixels. Leveraging the learned alignment, a classifier trained on one agent can be ported to the other with no additional gradient steps, while distillation-like migration accelerates later learning and markedly reduces total compute. The findings reveal that predictive learning objectives impose strong regularities on representation geometry, suggesting a lightweight path to interoperability among decentralized vision systems. The code is available at https://anonymous.4open.science/r/Social-JEPA-5C57.
Received — 25 February 2026 ⏭

MolReasoner: Toward Effective and Interpretable Reasoning for Molecular LLMs

✇cs.AI, q-bio.NC updates on arXiv.org
By: Guojiang Zhao · Zixiang Lu · Yutang Ge · Sihang Li · Zheng Cheng · Haitao Lin · Lirong Wu · Hanchen Xia · Hengxing Cai · Wentao Guo · Hongshuai Wang · Mingjun Xu · Siyu Zhu · Guolin Ke · Linfeng Zhang · Zhifeng Gao
24 February 2026 at 13:00
arXiv:2508.02066v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have shown impressive performance across various domains, but their ability to perform molecular reasoning remains underexplored. Existing methods mostly rely on general-purpose prompting, which lacks domain-specific molecular semantics, or fine-tuning, which faces challenges in interpretability and reasoning depth, often leading to structural and textual hallucinations. To address these issues, we introduce MolReasoner, a two-stage framework that transitions LLMs from memorization to high-fidelity chemical reasoning. In the Mol-SFT stage, knowledge-enhanced Chain-of-Thought (CoT) data provides a strong foundation, while the Mol-RL stage refines reasoning using a novel, task-adaptive reward system to mitigate hallucinations. Extensive evaluations demonstrate that MolReasoner significantly outperforms a wide range of strong baselines in both molecule generation and captioning tasks. Further analyses highlight the framework's synergistic design and its ability to produce more interpretable outputs. Our work presents a principled and effective new approach for advancing high-fidelity molecular reasoning.
Received — 17 February 2026 ⏭

LQA: A Lightweight Quantized-Adaptive Framework for Vision-Language Models on the Edge

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xin Wang · Hualin Zhou · Sheng Guang Wang · Ting Dang · Yu Zhang · Hong Jia · Tao Gu
17 February 2026 at 13:00
arXiv:2602.07849v2 Announce Type: replace Abstract: Deploying Vision-Language Models (VLMs) on edge devices is challenged by resource constraints and performance degradation under distribution shifts. While test-time adaptation (TTA) can counteract such shifts, existing methods are too resource-intensive for on-device deployment. To address this challenge, we propose LQA, a lightweight, quantized-adaptive framework for VLMs that combines a modality-aware quantization strategy with gradient-free test-time adaptation. We introduce Selective Hybrid Quantization (SHQ) and a quantized, gradient-free adaptation mechanism to enable robust and efficient VLM deployment on resource-constrained hardware. Experiments across both synthetic and real-world distribution shifts show that LQA improves overall adaptation performance by 4.5\%, uses less memory than full-precision models, and significantly outperforms gradient-based TTA methods, achieving up to 19.9$\times$ lower memory usage across seven open-source datasets. These results demonstrate that LQA offers a practical pathway for robust, privacy-preserving, and efficient VLM deployment on edge devices.
There are no more articles
❌