Explore the vast range of titles available.

Recently, self-supervised learning methods have been shown to be very powerful and efficient for yielding robust representation learning by maximizing the similarity across different augmented views in embedding vector space. However, the main challenge is generating different views with random cropping; the semantic feature might exist differently across different views leading to inappropriately maximizing similarity objective. We tackle this problem by introducing Heuristic Attention Representation Learning (HARL). This self-supervised framework relies on the joint embedding architecture in which the two neural networks are trained to produce similar embedding for different augmented views of the same image. HARL framework adopts prior visual object-level attention by generating a heuristic mask proposal for each training image and maximizes the abstract object-level embedding on vector space instead of whole image representation from previous works. As a result, HARL extracts the quality semantic representation from each training sample and outperforms existing self-supervised baselines on several downstream tasks. In addition, we provide efficient techniques based on conventional computer vision and deep learning methods for generating heuristic mask proposals on natural image datasets. Our HARL achieves +1.3% advancement in the ImageNet semi-supervised learning benchmark and +0.9% improvement in AP50 of the COCO object detection task over the previous state-of-the-art method BYOL. Our code implementation is available for both TensorFlow and PyTorch frameworks.

Recent self-supervised contrastive learning methods are powerful and efficient for robust representation learning, pulling semantic features from different cropping views of the same image while pushing other features away from other images in the embedding vector space. However, model training for contrastive learning is quite inefficient. In the high-dimensional vector space of the images, images can differ from each other in many ways. We address this problem with heuristic attention pixel-level contrastive loss for representation learning (HAPiCLR), a self-supervised joint embedding contrastive framework that operates at the pixel level and makes use of heuristic mask information. HAPiCLR leverages pixel-level information from the object’s contextual representation instead of identifying pair-wise differences in instance-level representations. Thus, HAPiCLR enhances contrastive learning objectives without requiring large batch sizes, memory banks, or queues, thereby reducing the memory footprint and the processing needed for large datasets. Furthermore, HAPiCLR loss combined with other contrastive objectives such as SimCLR or MoCo loss produces considerable performance boosts on all downstream tasks, including image classification, object detection, and instance segmentation.

In the real world, the appearance of identical objects depends on factors as varied as resolution, angle, illumination conditions, and viewing perspectives. This suggests that the data augmentation pipeline could benefit downstream tasks by exploring the overall data appearance in a self-supervised framework. Previous work on self-supervised learning that yields outstanding performance relies heavily on data augmentation such as cropping and color distortion. However, most methods use a static data augmentation pipeline, limiting the amount of feature exploration. To generate representations that encompass scale-invariant, explicit information about various semantic features and are invariant to nuisance factors such as relative object location, brightness, and color distortion, we propose the Multi-View, Multi-Augmentation (MVMA) framework. MVMA consists of multiple augmentation pipelines, with each pipeline comprising an assortment of augmentation policies. By refining the baseline self-supervised framework to investigate a broader range of image appearances through modified loss objective functions, MVMA enhances the exploration of image features through diverse data augmentation techniques. Transferring the resultant representation learning using convolutional networks (ConvNets) to downstream tasks yields significant improvements compared to the state-of-the-art DINO across a wide range of vision tasks and classification tasks: +4.1% and +8.8% top-1 on the ImageNet dataset with linear evaluation and k-NN classifier, respectively. Moreover, MVMA achieves a significant improvement of +5% AP50 and +7% AP50m on COCO object detection and segmentation.

Recent advances in robot world models enable synthetic video generation for embodied prediction and planning. However, evaluating these videos is challenging: visually realistic outputs often violate physical laws, temporal consistency, or task logic, while conventional metrics and monolithic Vision-Language Model (VLM) judges fail to generalize or provide precise diagnostic value. We present RoboGaze, a training-free, multi-agent VLM framework that provides structured, interpretable evaluation for generated robot-manipulation videos. Given a task instruction and video, RoboGaze operates via a three-stage pipeline: task-scene grounding, dimension-specific specialist routing, and critic-based verification. It outputs temporally localized glitch reports categorized under a novel 6-dimension, 30-type robotics-specific taxonomy. To benchmark RoboGaze, we introduce a human-validated dataset of 382 clips spanning simulated and real-world multi-view manipulation. Evaluating eight open-source and proprietary VLM backbones, RoboGaze dramatically outperforms zero-shot baselines, improving description-F1 by up to +43 points and temporal alignment (F1 x IoU) by up to +37 points, closing approximately 85% of the gap to the human ceiling. Furthermore, its critic verifier mitigates the \"cry-wolf\" false-positive flaw of standard VLMs, lifting clean-clip accuracy from under 25% to over 80%. RoboGaze offers a scalable, highly interpretable diagnostic tool for the rigorous evaluation of robot world models.

Affordance understanding, the task of identifying actionable regions on 3D objects, plays a vital role in allowing robotic systems to engage with and operate within the physical world. Although Visual Language Models (VLMs) have excelled in high-level reasoning and long-horizon planning for robotic manipulation, they still fall short in grasping the nuanced physical properties required for effective human-robot interaction. In this paper, we introduce PAVLM (Point cloud Affordance Vision-Language Model), an innovative framework that utilizes the extensive multimodal knowledge embedded in pre-trained language models to enhance 3D affordance understanding of point cloud. PAVLM integrates a geometric-guided propagation module with hidden embeddings from large language models (LLMs) to enrich visual semantics. On the language side, we prompt Llama-3.1 models to generate refined context-aware text, augmenting the instructional input with deeper semantic cues. Experimental results on the 3D-AffordanceNet benchmark demonstrate that PAVLM outperforms baseline methods for both full and partial point clouds, particularly excelling in its generalization to novel open-world affordance tasks of 3D objects. For more information, visit our project site: pavlm-source.github.io.

Affordance understanding, the task of identifying actionable regions on 3D objects, plays a vital role in allowing robotic systems to engage with and operate within the physical world. Although Visual Language Models (VLMs) have excelled in high-level reasoning and long-horizon planning for robotic manipulation, they still fall short in grasping the nuanced physical properties required for effective human-robot interaction. In this paper, we introduce PAVLM (Point cloud Affordance Vision-Language Model), an innovative framework that utilizes the extensive multimodal knowledge embedded in pre-trained language models to enhance 3D affordance understanding of point cloud. PAVLM integrates a geometric-guided propagation module with hidden embeddings from large language models (LLMs) to enrich visual semantics. On the language side, we prompt Llama-3.1 models to generate refined context-aware text, augmenting the instructional input with deeper semantic cues. Experimental results on the 3D-AffordanceNet benchmark demonstrate that PAVLM outperforms baseline methods for both full and partial point clouds, particularly excelling in its generalization to novel open-world affordance tasks of 3D objects. For more information, visit our project site: pavlm-source.github.io.

In recent years, self-supervised learning has been studied to deal with the limitation of available labeled-dataset. Among the major components of self-supervised learning, the data augmentation pipeline is one key factor in enhancing the resulting performance. However, most researchers manually designed the augmentation pipeline, and the limited collections of transformation may cause the lack of robustness of the learned feature representation. In this work, we proposed Multi-Augmentations for Self-Supervised Representation Learning (MA-SSRL), which fully searched for various augmentation policies to build the entire pipeline to improve the robustness of the learned feature representation. MA-SSRL successfully learns the invariant feature representation and presents an efficient, effective, and adaptable data augmentation pipeline for self-supervised pre-training on different distribution and domain datasets. MA-SSRL outperforms the previous state-of-the-art methods on transfer and semi-supervised benchmarks while requiring fewer training epochs.

Five new secondary metabolites, modiolides D-G (1−4) and 1-(2,5-dihydroxyphenyl)-3-methoxy-butan-1-one (8), one new natural product, 1-(2,5-dihydroxyphenyl)-3-hydroxybutan-1-one (7), along with three known compounds, modiolides A (5) and B (6), and 1-(2,5-dihydroxyphenyl)-2-buten-1-one (9) were isolated from a fermentation culture of the marine endophytic fungus Paraconiothyrium sp. VK-13. Their chemical structures were elucidated by the NMR and MS spectroscopic analysis as well as the modified Mosher’s method. Compounds 7 and 9 inhibited the overproduction of proinflammatory mediators NO and PGE2 in LPS-stimulated RAW264.7 cells, with IC50 values ranging from 3.9 to 12.5 µM. The inhibitory effects of 7 and 9 on the release of NO and PGE2 were correlated with their significant suppression of iNOS and COX-2 protein expression, respectively. Furthermore, both compounds 7 and 9 inhibited the mRNA expression of proinflammatory cytokines, including TNF-α, IL-1β, IL-6, and IL-12, with IC50 values in a range of 2.4−12.5 µM.

Chemical investigation of the Kappaphycus alvarezii-derived endophytic fungus Aspergillus micronesiensis lead to the isolation of three novel dibenzospiroketals, aspermicrones A-C (1-3). Their chemical structures were determined by extensive analysis of HR-ESI-MS and NMR spectral data. The absolute configurations of them were determined by experimental and TD-DFT theoretical calculated circular dichroism spectra. Compound 2 exhibited selective cytotoxic effect toward HepG2 cell line (IC50 = 9.9 µM). Additionally, both of compounds 2 and 3 displayed anti-microbial activity against Staphylococcus aureus (MIC = 123.2 µM for each compound). Compound 1 was inactivity in both cytotoxic and anti-microbial assays.

Context:Paramignya trimera (Oliv.) Burkill (Rutaceae) has been used to treat liver diseases and cancer. However, the anti-inflammatory effects of this medicinal plant and its components have not been elucidated.Objective: This study investigated chemical constituents of the P. trimera stems and evaluated anti-inflammatory effects of isolated compounds.Materials and methods: Cytotoxicity of isolated compounds (5–40 μM) toward BV2 cells was tested using 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) for 24 h. Inhibitory effects of isolated compounds (5-40 μM) on nitrite and PGE2 concentrations were determined using Griess reaction and PGE2 ELISA kit, respectively (pretreated with the compounds for 3 h and then stimulated for 18 h with LPS). Inhibitory effects of compounds (5-40 μM) on iNOS and COX-2 protein expression were evaluated by Western blot analysis (pretreated with the compounds for 3 h and then stimulated for 24 h with LPS).Results: Seven coumarins were isolated and identified as: ostruthin (1), ninhvanin (2), 8-geranyl-7-hydroxycoumarin (3), 6-(6′,7′-dihydroxy-3′,7′-dimethylocta-2′-enyl)-7-hydroxycoumarin (4), 6-(7-hydroperoxy-3,7-dimethylocta-2,5-dienyl)-7-hydroxycoumarin (5), 6-(2-hydroxyethyl)-2,2-dimethyl-2H-1-benzopyran (6), and luvangetin (7). Compounds 1–4 and 7 inhibited NO and PGE2 production in LPS-stimulated BV2 cells, with IC50 values ranging from 9.8 to 46.8 and from 9.4 to 52.8 μM, respectively. Ostruthin (1) and ninhvanin (2) were shown to suppress LPS-induced iNOS and COX-2 protein expression.Discussion and conclusion: The present study provides a scientific rationale for the use of P. trimera in the prevention and treatment of neuroinflammatory diseases. Ostruthin and ninhvanin might have potential therapeutic effects and should be considered for further development as new anti-neuroinflammatory agents.