NSTL国家科技图书文献中心

1. IMC-Det: Intra-Inter Modality Contrastive Learning for Video Object Detection NSTL国家科技图书文献中心

Qi, Qiang | Qiu, Zhenyu ... - 《International Journal of Computer Vision》 - 2025,133(2) - 890～909 - 共20页

摘要： inter-modality contrastive learning module to enforce | knowledge of the textual modality due to the large inter | performance. In this paper, we propose a novel intra-inter | modality contrastive learning network for high | . First, we design an intra-modality* contrastive*

关键词： Video object detection | Intra-inter modality learning | Contrastive learning | Feature aggregation

2. DMANet: Dual-modality alignment network for visible-infrared person re-identification NSTL国家科技图书文献中心

Cheng, Xu | Deng, Shuya ... - 《Pattern Recognition》 - 2025,157 - 共10页

摘要： to align well due to the large inter- and intra | solve modality* difference by proposing inter- and* | intra- modality alignment module (IIMA). Specifically | intra- modality alignment module is presented to | intra- modalities. Finally, joint learning mechanism

关键词： Visible-infrared person re-identification | Transformer | Attention mechanism | Modality alignment

3. Bridging the Modality Gap in Multimodal Eye Disease Screening: Learning Modality Shared-Specific Features via Multi-Level Regularization NSTL国家科技图书文献中心

Jiayue Zhao | Shiman Li ... - 《IEEE signal processing letters》 - 2025,32 - 586～590 - 共5页

摘要： regularization strategy including intra-modality, inter | -modality, and intra-inter-modality* regularization to* | facilitate the effective learning* of the modality Shared* | . However, existing multimodal learning* methods cannot* | fully leverage the information from each modality* due*

关键词： Feature extraction | Translation | Three-dimensional displays | Semantics | Vectors | Bridges | Glaucoma | Data mining | Transforms | Training

4. I2OL-Net: Intra-Inter Objectness Learning Network for Point-Supervised X-Ray Prohibited Item Detection NSTL国家科技图书文献中心

Chenyang Wang | Yan Yan ... - 《IEEE transactions on information forensics and security》 - 2025,20 - 3045～3059 - 共15页

摘要： learning (intra-OL) module and an inter-modality | objectness learning* (inter-OL) module. The intra-OL module* | -efficient point supervision and develop an intra-inter | of two key modules: an intra-modality* objectness* | objectness learning* network (I2OL-Net). I2OL-Net consists*

关键词： X-ray imaging | Annotations | Object detection | Proposals | Detectors | Costs | Adversarial machine learning | Reviews | Training | Supervised learning

5. Image-text feature learning for unsupervised visible-infrared person re-identification NSTL国家科技图书文献中心

Guo, Jifeng | Pang, Zhiqi - 《Image and vision computing》 - 2025,158(May) - 1.1～1.9 - 共9页

摘要： intra-modality identity-related learning but also | effectiveness of text semantics in inter-modality* matching and* | modality-invariant feature learning. In this paper, we | incorporates text semantics into inter-modality* matching and* | modality-invariant feature learning. Specifically, ITFL

关键词： Unsupervised learning | Visible-infrared person re-identification | Contrastive learning | Vision-language models

6. Decoupling Common and Unique Representations for Multimodal Self-supervised Learning NSTL国家科技图书文献中心

Yi Wang | Conrad M. Albrecht ... - 《Computer Vision - ECCV 2024,Part XXIX》 - European Conference on Computer Vision - 2025, - 286～303 - 共18页

摘要： while ignoring intra-modal training and modality | . By distinguishing inter- and intra-modal embeddings | -supervised learning. However, most existing approaches | effective method for multimodal self-supervised learning | modality-missing settings. With thorough experiments and

关键词： Self-supervised learning | Multimodal representations

7. Learning Modality-Agnostic Representation for Semantic Segmentation from Any Modalities NSTL国家科技图书文献中心

Xu Zheng | Yuanhuiyi Lyu ... - 《Computer Vision - ECCV 2024,Part XXXIV》 - European Conference on Computer Vision - 2025, - 146～165 - 共20页

摘要： direction in learning* the modality-agnostic representation* | Image modality* is not perfect as it often* | segmentation methods when confronting modality* absence or* | . Inspired by the open-world learning* capability of multi* | distillation (LSCD) module to transfer both inter-modal and

关键词： Multi-modal segmentation | MVLMs | KD

8. Multi-Sensor Learning Enables Information Transfer Across Different Sensory Data and Augments Multi-Modality Imaging NSTL国家科技图书文献中心

Lingting Zhu | Yizheng Chen ... - 《IEEE Transactions on Pattern Analysis and Machine Intelligence》 - 2025,47(1) - 288～304 - 共17页 - 被引量：1

摘要： features in multi-modality* imaging, namely intra- and* | inter-modality features, and present a multi-sensor | learning (MSL) framework to utilize the crossover inter | Multi-modality* imaging is widely used in* | , multi-modality* imaging is accomplished by post hoc*

关键词： Magnetic resonance imaging | Computed tomography | Imaging | Feature extraction | Transformers | Image reconstruction | Biomedical imaging

9. I2M2Net: Inter/Intra-modal Feature Masking Self-distillation for Incomplete Multimodal Skin Lesion Diagnosis NSTL国家科技图书文献中心

Ke Wang | Linwei Qiu ... - 《Medical Image Computing and Computer Assisted Intervention - MICCAI 2024 Workshops》 - International Conference on Medical Image Computing and Computer Assisted Intervention | International Workshop on Interpretability of Machine Intelligence in Medical Image Computing | Workshop on Distributed,Collaborative,and Federated Learning | Embodied AI and Robotics for HealTHcare | International Skin Imaging Collaboration - 2025, - 3～13 - 共11页

摘要：) random dropout of modality* (i.e. inter-modal feature* | Multimodal learning* has demonstrated promising* | learning, focusing on adaptively and progressively mining | to perform self-distillation with masked modality | masking) to simulate different missing modality

关键词： Missing modality | Multi-modal learning | Skin lesion diagnosis

10. UAE: Universal Anatomical Embedding on multi-modality medical images NSTL国家科技图书文献中心

Bai, Xiaoyu | Bai, Fan ... - 《Medical image analysis》 - 2025,103 - 103562～103562 - 共13页

摘要： intra-and inter-modality tasks, including one-shot | self-supervised learning* to create a discriminative* | (3) cross-modality* matching (e.g., CT-MRI landmark* | propose a Unified framework for learning* Anatomical* | , semantic, and cross-modality* anatomical embeddings*

摘要： inter-modality contrastive learning module to enforce | knowledge of the textual modality due to the large inter | performance. In this paper, we propose a novel intra-inter | modality contrastive learning network for high | . First, we design an intra-modality* contrastive*

关键词： Video object detection | Intra-inter modality learning | Contrastive learning | Feature aggregation

摘要： to align well due to the large inter- and intra | solve modality* difference by proposing inter- and* | intra- modality alignment module (IIMA). Specifically | intra- modality alignment module is presented to | intra- modalities. Finally, joint learning mechanism

关键词： Visible-infrared person re-identification | Transformer | Attention mechanism | Modality alignment

关键词： Feature extraction | Translation | Three-dimensional displays | Semantics | Vectors | Bridges | Glaucoma | Data mining | Transforms | Training

摘要： learning (intra-OL) module and an inter-modality | objectness learning* (inter-OL) module. The intra-OL module* | -efficient point supervision and develop an intra-inter | of two key modules: an intra-modality* objectness* | objectness learning* network (I2OL-Net). I2OL-Net consists*

关键词： X-ray imaging | Annotations | Object detection | Proposals | Detectors | Costs | Adversarial machine learning | Reviews | Training | Supervised learning

关键词： Unsupervised learning | Visible-infrared person re-identification | Contrastive learning | Vision-language models

摘要： while ignoring intra-modal training and modality | . By distinguishing inter- and intra-modal embeddings | -supervised learning. However, most existing approaches | effective method for multimodal self-supervised learning | modality-missing settings. With thorough experiments and

关键词： Self-supervised learning | Multimodal representations

关键词： Multi-modal segmentation | MVLMs | KD

摘要： features in multi-modality* imaging, namely intra- and* | inter-modality features, and present a multi-sensor | learning (MSL) framework to utilize the crossover inter | Multi-modality* imaging is widely used in* | , multi-modality* imaging is accomplished by post hoc*

关键词： Magnetic resonance imaging | Computed tomography | Imaging | Feature extraction | Transformers | Image reconstruction | Biomedical imaging

摘要：) random dropout of modality* (i.e. inter-modal feature* | Multimodal learning* has demonstrated promising* | learning, focusing on adaptively and progressively mining | to perform self-distillation with masked modality | masking) to simulate different missing modality

关键词： Missing modality | Multi-modal learning | Skin lesion diagnosis

摘要： intra-and inter-modality tasks, including one-shot | self-supervised learning* to create a discriminative* | (3) cross-modality* matching (e.g., CT-MRI landmark* | propose a Unified framework for learning* Anatomical* | , semantic, and cross-modality* anatomical embeddings*

关键词： Anatomical embedding learning | Landmark matching | Multi-modality image alignment

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： inter-modality contrastive learning module to enforce | knowledge of the textual modality due to the large inter | performance. In this paper, we propose a novel intra-inter | modality contrastive learning network for high | . First, we design an intra-modality contrastive

关键词： Video object detection | Intra-inter modality learning | Contrastive learning | Feature aggregation

摘要： to align well due to the large inter- and intra | solve modality difference by proposing inter- and | intra- modality alignment module (IIMA). Specifically | intra- modality alignment module is presented to | intra- modalities. Finally, joint learning mechanism

关键词： Visible-infrared person re-identification | Transformer | Attention mechanism | Modality alignment

摘要： regularization strategy including intra-modality, inter | -modality, and intra-inter-modality regularization to | facilitate the effective learning of the modality Shared | . However, existing multimodal learning methods cannot | fully leverage the information from each modality due

关键词： Feature extraction | Translation | Three-dimensional displays | Semantics | Vectors | Bridges | Glaucoma | Data mining | Transforms | Training

摘要： learning (intra-OL) module and an inter-modality | objectness learning (inter-OL) module. The intra-OL module | -efficient point supervision and develop an intra-inter | of two key modules: an intra-modality objectness | objectness learning network (I2OL-Net). I2OL-Net consists

关键词： X-ray imaging | Annotations | Object detection | Proposals | Detectors | Costs | Adversarial machine learning | Reviews | Training | Supervised learning

关键词： Unsupervised learning | Visible-infrared person re-identification | Contrastive learning | Vision-language models

摘要： while ignoring intra-modal training and modality | . By distinguishing inter- and intra-modal embeddings | -supervised learning. However, most existing approaches | effective method for multimodal self-supervised learning | modality-missing settings. With thorough experiments and

关键词： Self-supervised learning | Multimodal representations

摘要： direction in learning the modality-agnostic representation | Image modality is not perfect as it often | segmentation methods when confronting modality absence or | . Inspired by the open-world learning capability of multi | distillation (LSCD) module to transfer both inter-modal and

关键词： Multi-modal segmentation | MVLMs | KD

摘要： features in multi-modality imaging, namely intra- and | inter-modality features, and present a multi-sensor | learning (MSL) framework to utilize the crossover inter | Multi-modality imaging is widely used in | , multi-modality imaging is accomplished by post hoc

关键词： Magnetic resonance imaging | Computed tomography | Imaging | Feature extraction | Transformers | Image reconstruction | Biomedical imaging

摘要：) random dropout of modality (i.e. inter-modal feature | Multimodal learning has demonstrated promising | learning, focusing on adaptively and progressively mining | to perform self-distillation with masked modality | masking) to simulate different missing modality

关键词： Missing modality | Multi-modal learning | Skin lesion diagnosis

摘要： intra-and inter-modality tasks, including one-shot | self-supervised learning to create a discriminative | (3) cross-modality matching (e.g., CT-MRI landmark | propose a Unified framework for learning Anatomical | , semantic, and cross-modality anatomical embeddings

关键词： Anatomical embedding learning | Landmark matching | Multi-modality image alignment

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： inter-modality contrastive learning module to enforce | knowledge of the textual modality due to the large inter | performance. In this paper, we propose a novel intra-inter | modality contrastive learning network for high | . First, we design an intra-modality* contrastive*

摘要： to align well due to the large inter- and intra | solve modality* difference by proposing inter- and* | intra- modality alignment module (IIMA). Specifically | intra- modality alignment module is presented to | intra- modalities. Finally, joint learning mechanism

摘要： learning (intra-OL) module and an inter-modality | objectness learning* (inter-OL) module. The intra-OL module* | -efficient point supervision and develop an intra-inter | of two key modules: an intra-modality* objectness* | objectness learning* network (I2OL-Net). I2OL-Net consists*

摘要： features in multi-modality* imaging, namely intra- and* | inter-modality features, and present a multi-sensor | learning (MSL) framework to utilize the crossover inter | Multi-modality* imaging is widely used in* | , multi-modality* imaging is accomplished by post hoc*

摘要：) random dropout of modality* (i.e. inter-modal feature* | Multimodal learning* has demonstrated promising* | learning, focusing on adaptively and progressively mining | to perform self-distillation with masked modality | masking) to simulate different missing modality

摘要： intra-and inter-modality tasks, including one-shot | self-supervised learning* to create a discriminative* | (3) cross-modality* matching (e.g., CT-MRI landmark* | propose a Unified framework for learning* Anatomical* | , semantic, and cross-modality* anatomical embeddings*