Traditional dialogue retrieval aims to select the most appropriate utterance or image from recent dialogue history. However, they often fail to meet users’ actual needs for revisiting semantically coherent content scattered across long-form conversations. To fill this gap, we define the Fine-grained Fragment Retrieval (FFR) task, requiring models to locate query-relevant fragments, comprising both utterances and images, from multimodal long-form dialogues. As a foundation for FFR, we construct MLDR, the longest-turn multimodal dialogue retrieval dataset to date, averaging 25.45 turns per dialogue, with each naturally spanning three distinct topics. To evaluate generalization in real-world scenarios, we curate and annotate a WeChat-based test set comprising real-world multimodal dialogues with an average of 75.38 turns. Building on these resources, we explore existing generation-based Vision-Language Models (VLMs) on FFR and observe that they often retrieve incoherent utterance-image fragments. While optimized for generating responses from visual-textual inputs, these models lack explicit supervision to ensure semantic coherence within retrieved fragments. To address this, we propose F2RVLM, a generative retrieval model trained in a two-stage paradigm: (1) supervised fine-tuning to inject fragment-level retrieval knowledge, and (2) GRPO-based reinforcement learning with multi-objective rewards to encourage outputs with semantic precision, relevance, and contextual coherence. In addition, to account for difficulty variations arising from differences in intra-fragment element distribution, ranging from locally dense to sparsely scattered, we introduce a difficulty-aware curriculum sampling that ranks training instances by predicted difficulty and gradually incorporates harder examples. This strategy enhances the model’s reasoning ability in long-form, multi-turn dialogue contexts. Experiments on both in-domain and real-domain sets demonstrate that F2RVLM substantially outperforms popular VLMs, achieving superior retrieval performance.

The rapid advancement of foundation models has revolutionized visual representation learning in a self-supervised manner. However, their application in remote sensing (RS) remains constrained by a fundamental gap: existing models predominantly handle single or limited modalities, overlooking the inherently multi-modal nature of RS observations. Optical, synthetic aperture radar (SAR), and multi-spectral data offer complementary insights that significantly reduce the inherent ambiguity and uncertainty in single-source analysis. To bridge this gap, we introduce RingMoE, a unified multi-modal RS foundation model with 14.7 billion parameters, pre-trained on 400 million multi-modal RS images from nine satellites. RingMoE incorporates three key innovations: 1) A hierarchical Mixture-of-Experts (MoE) architecture comprising modal-specialized, collaborative, and shared experts, effectively modeling intra-modal knowledge while capturing cross-modal dependencies to mitigate conflicts between modal representations; 2) Physics-informed self-supervised learning, explicitly embedding sensor-specific radiometric characteristics into the pre-training objectives; 3) Dynamic expert pruning, enabling adaptive model compression from 14.7B to 1B parameters while maintaining performance, facilitating efficient deployment in Earth observation applications. Evaluated across 23 benchmarks spanning six key RS tasks (i.e., classification, detection, segmentation, tracking, change detection, and depth estimation), RingMoE outperforms existing foundation models and sets new SOTAs, demonstrating remarkable adaptability from single-modal to multi-modal scenarios. Beyond theoretical progress, it has been deployed and trialed in multiple sectors, including emergency response, land management, marine sciences, and urban planning.

Vision foundation models in remote sensing have been extensively studied due to their superior generalization on various downstream tasks. Synthetic Aperture Radar (SAR) offers all-day, all-weather imaging capabilities, providing significant advantages for Earth observation. However, establishing a foundation model for SAR image interpretation inevitably encounters the challenges of insufficient information utilization and poor interpretability. In this paper, we propose a remote sensing foundation model based on complex-valued SAR data, which simulates the polarimetric decomposition process for pre-training, i.e., characterizing pixel scattering intensity as a weighted combination of scattering bases and scattering coefficients, thereby endowing the foundation model with physical interpretability. Specifically, we construct a series of scattering queries, each representing an independent and meaningful scattering basis, which interact with SAR features in the scattering query decoder and output the corresponding scattering coefficient. To guide the pre-training process, polarimetric decomposition loss and power self-supervised loss are constructed. The former aligns the predicted coefficients with Yamaguchi coefficients, while the latter reconstructs power from the predicted coefficients and compares it to the input image's power. The performance of our foundation model is validated on nine typical downstream tasks, achieving state-of-the-art results. Notably, the foundation model can extract stable feature representations and exhibits strong generalization, even in data-scarce conditions.

Remote sensing foundation models largely break away from the traditional paradigm of designing task-specific models, offering greater scalability across multiple tasks. However, they face challenges such as low computational efficiency and limited interpretability, especially when dealing with large-scale remote sensing images. To overcome these, we draw inspiration from heat conduction, a physical process modeling local heat diffusion. Building on this idea, we are the first to explore the potential of using the parallel computing model of heat conduction to simulate the local region correlations in high-resolution remote sensing images, and introduce RS-vHeat, an efficient multi-modal remote sensing foundation model. Specifically, RS-vHeat 1) applies the Heat Conduction Operator (HCO) with a complexity of $O(N^{1.5})$ and a global receptive field, reducing computational overhead while capturing remote sensing object structure information to guide heat diffusion; 2) learns the frequency distribution representations of various scenes through a self-supervised strategy based on frequency domain hierarchical masking and multi-domain reconstruction; 3) significantly improves efficiency and performance over state-of-the-art techniques across 4 tasks and 10 datasets. Compared to attention-based remote sensing foundation models, we reduce memory usage by 84\%, FLOPs by 24\% and improves throughput by 2.7 times. The code will be made publicly available.

For more efficient generalization to unseen domains (classes), most Few-shot Segmentation (FSS) would directly exploit pre-trained encoders and only fine-tune the decoder, especially in the current era of large models. However, such fixed feature encoders tend to be class-agnostic, inevitably activating objects that are irrelevant to the target class. In contrast, humans can effortlessly focus on specific objects in the line of sight. This paper mimics the visual perception pattern of human beings and proposes a novel and powerful prompt-driven scheme, called "Prompt and Transfer" (PAT), which constructs a dynamic class-aware prompting paradigm to tune the encoder for focusing on the interested object (target class) in the current task. Three key points are elaborated to enhance the prompting: 1) Cross-modal linguistic information is introduced to initialize prompts for each task. 2) Semantic Prompt Transfer (SPT) that precisely transfers the class-specific semantics within the images to prompts. 3) Part Mask Generator (PMG) that works in conjunction with SPT to adaptively generate different but complementary part prompts for different individuals. Surprisingly, PAT achieves competitive performance on 4 different tasks including standard FSS, Cross-domain FSS (e.g., CV, medical, and remote sensing domains), Weak-label FSS, and Zero-shot Segmentation, setting new state-of-the-arts on 11 benchmarks.

Aerial Remote Sensing (ARS) vision tasks present significant challenges due to the unique viewing angle characteristics. Existing research has primarily focused on algorithms for specific tasks, which have limited applicability in a broad range of ARS vision applications. This paper proposes RingMo-Aerial, aiming to fill the gap in foundation model research in the field of ARS vision. A Frequency-Enhanced Multi-Head Self-Attention (FE-MSA) mechanism is introduced to strengthen the model's capacity for small-object representation. Complementarily, an affine transformation-based contrastive learning method improves its adaptability to the tilted viewing angles inherent in ARS tasks. Furthermore, the ARS-Adapter, an efficient parameter fine-tuning method, is proposed to improve the model's adaptability and performance in various ARS vision tasks. Experimental results demonstrate that RingMo-Aerial achieves SOTA performance on multiple downstream tasks. This indicates the practicality and efficacy of RingMo-Aerial in enhancing the performance of ARS vision tasks.

The remote sensing visual grounding (RSVG) task focuses on accurately identifying and localizing specific targets in remote sensing (RS) images using descriptive query expressions. Existing methods independently extract visual and textual features, ignoring early complementary information between image and text. This leads to information loss and misalignment, limiting the model’s ability to distinguish similar targets. To address this challenge, we propose the query-aware multimodal fusion network (QAMFN), which introduces an innovative query-guided visual attention (QGVA) mechanism in the early stages of the visual encoder. This mechanism integrates textual information during the early visual feature extraction process, thereby resolving the issue of missing image-text complementary information. QGVA ensures that the visual backbone accurately focuses on local features highly relevant to the query by injecting textual information into the visual encoding process. Additionally, to enhance the model’s ability to integrate multimodal information and adapt to more complex RS images, we introduce the text-semantic attention-guided masking (TAM) module. TAM aggregates multimodal features processed by the backbones and filters out redundant information, producing high-quality fused features. Experiments demonstrate that our approach sets a new record on the DIOR-RSVG dataset, improving accuracy to 81.67% (an absolute increase of 4.98%).

Research on multiview stereo (MVS) based on remote sensing images has promoted the development of large-scale urban 3-D reconstruction. However, remote sensing multiview image data suffer from the problems of occlusion and uneven brightness between views during acquisition, which leads to the problem of blurred details in depth estimation. To solve the above problem, we reexamine the deformable learning method in the MVS task and propose a novel paradigm based on view space and depth deformable learning (SDL-MVS), aiming to learn deformable interactions of features in different view spaces and deformably model the depth ranges and intervals to enable high accurate depth estimation. Specifically, to solve the problem of view noise caused by occlusion and uneven brightness, we propose a progressive space deformable sampling (PSS) mechanism, which performs deformable learning of sampling points in the 3-D frustum space and the 2-D image space in a progressive manner to embed source features to the reference feature adaptively. To further optimize the depth, we introduce depth hypothesis deformable discretization (DHD), which achieves precise positioning of the depth prior by adaptively adjusting the depth range hypothesis and performing deformable discretization of the depth interval hypothesis. Finally, our SDL-MVS achieves explicit modeling of occlusion and uneven brightness faced in MVS through the deformable learning paradigm of view space and depth, achieving accurate multiview depth estimation. Extensive experiments on LuoJia-MVS and WHU datasets show that our SDL-MVS reaches state-of-the-art performance. It is worth noting that our SDL-MVS achieves a mean absolute error (MAE) error of 0.086 and an accuracy of 98.9% for Acc$_{\lt 0.6\,\text {m}}$and 98.9% for Acc$_{\lt 3-\text {interval}}$on the LuoJia-MVS dataset under the premise of three views as input.

Few-shot remote sensing image classification entails identifying images using a limited set of labeled data within remote sensing scenes, holding significant theoretical and practical implications. However, owing to the intricacy and variety of remote sensing images, traditional classification methods usually struggle to extract effective features and learn robust classifiers. To address this issue, an end-to-end metric learning framework named Attention-based Contrastive Learning Network is introduced in this paper. Specifically, the Attention-based Feature Optimization (ABFO) module is employed to align and enhance target image features, highlighting the target region and strengthening the network’s feature extraction capability. Additionally, the Dictionary-based Contrastive Loss (DBCL) module is assigned to optimize image feature vectors, improving category distinguishability and consequently enhancing classification accuracy. The experimental results on five publicly available Few-shot remote sensing classification datasets demonstrate the high competitiveness of our proposed method. Furthermore, it illustrates superior classification accuracy compared to other pertinent Few-shot learning algorithms in the 5-way 1-shot scenario.

As drone technology advances, using unmanned aerial vehicles for aerial surveys has become the dominant trend in modern low-altitude remote sensing. The surge in aerial video data necessitates accurate prediction for future scenarios and motion states of the interested target, particularly in applications like traffic management and disaster response. Existing video prediction methods focus solely on predicting future scenes (video frames), suffering from the neglect of explicitly modeling target’s motion states, which is crucial for aerial video interpretation. To address this issue, we introduce a novel task called Target-Aware Aerial Video Prediction, aiming to simultaneously predict future scenes and motion states of the target. Further, we design a model specifically for this task, named TAFormer, which provides a unified modeling approach for both video and target motion states. Specifically, we introduce Spatiotemporal Attention (STA), which decouples the learning of video dynamics into spatial static attention and temporal dynamic attention, effectively modeling the scene appearance and motion. Additionally, we design an Information Sharing Mechanism (ISM), which elegantly unifies the modeling of video and target motion by facilitating information interaction through two sets of messenger tokens. Moreover, to alleviate the difficulty of distinguishing targets in blurry predictions, we introduce Target-Sensitive Gaussian Loss (TSGL), enhancing the model’s sensitivity to both target’s position and content. Extensive experiments on UAV123VP and VisDroneVP (derived from single-object tracking datasets) demonstrate the exceptional performance of TAFormer in target-aware video prediction, showcasing its adaptability to the additional requirements of aerial video interpretation for target awareness.

The construction of a basic model to extract generalized features from a large number of multimodal data is a new challenge in the field of remote sensing. Compared with natural scene images, When faced with a complex application scenario of remote sensing of multi-sensor acquisition, models that are suitable for a specific task are difficult to generalize to new scenarios. In this paper, we propose a model architecture based on the concepts of multi-domain representation and cross-domain fusion. By extracting strong generalization features from massive multi-modal data, a single foundation model can accomplish generalization interpretation for multiple downstream tasks. Experimental results show that the proposed model performs well on multiple downstream tasks, which validates the feasibility of the remote sensing cross-modal foundation model in the interpretation task.

Few-shot segmentation (FSS) is proposed to segment unknown class targets with just a few annotated samples. Most current FSS methods follow the paradigm of mining the semantics from the support images to guide the query image segmentation. However, such a pattern of ‘learning from others’ struggles to handle the extreme intra-class variation, preventing FSS from being directly generalized to remote sensing scenes. To bridge the gap of intra-class variance, we develop a Dual-Mining network named DMNet for cross-image mining and self-mining, meaning that it no longer focuses solely on support images but pays more attention to the query image itself. Specifically, we propose a Class-public Region Mining (CPRM) module to effectively suppress irrelevant feature pollution by capturing the common semantics between the support-query image pair. The Class-specific Region Mining (CSRM) module is then proposed to continuously mine the class-specific semantics of the query image itself in a ‘filtering’ and ‘purifying’ manner. In addition, to prevent the co-existence of multiple classes in remote sensing scenes from exacerbating the collapse of FSS generalization, we also propose a new Known-class Meta Suppressor (KMS) module to suppress the activation of known-class objects in the sample. Extensive experiments on the iSAID and LoveDA remote sensing datasets have demonstrated that our method sets the state-of-the-art with a minimum number of model parameters. Significantly, our model with the backbone of Resnet-50 achieves the mIoU of 49.58% and 51.34% on iSAID under 1-shot and 5-shot settings, outperforming the state-of-the-art method by 1.8% and 1.12%, respectively. The code is publicly available at https://github.com/HanboBizl/DMNet/.