Recent advances in retrieval-based in-context learning (ICL) train the retriever using a classification objective, which categorizes in-context examples (ICEs) into the most useful and the rest based on absolute scores. However, during inference, ICEs are retrieved by score ranking rather than classification -- The classification training objective deviates from this test scenario. Hence, in this paper, we propose a novel algorithm that trains a retrieval model by ranking formulation, where the preference rankings between ICEs are given by comparing the likelihood of the LLM generating the correct answer conditioned on each exemplar. By learning to rank, we motivate the retriever to automatically learn diverse rationales why specific examples are more useful for ICL decisions. This addresses the issue that classification models poorly capture broader utility. Experimental results demonstrate the top-1 performance of our proposal across 9 NLP tasks, with ablation studies and case studies further validating the effectiveness of our design.

Recent advances in multi-modal representation learning have led to unified embedding spaces that align modalities such as images, text, audio, and vision. However, human motion sequences, a modality that is fundamental for understanding dynamic human activities, remains largely unrepresented in these frameworks. Semantic understanding of actions requires multi-modal grounding: text conveys descriptive semantics, vision provides visual context, and audio provides environmental cues. To bridge this gap, we propose MotionBind, a novel architecture that extends the LanguageBind embedding space to incorporate human motion. MotionBind has two major components. The first one is a Multi-Scale Temporal Motion Transformer (MuTMoT) that maps motion sequences to semantically meaningful embeddings. Multimodal alignment is achieved via diverse cross-modal supervision, including motion-text pairs from HumanML3D and KIT-ML, motion-video pairs rendered from AMASS, and motion-video-audio triplets from AIST++. The second component is a Retrieval-Augmented Latent diffusion Model (REALM) that can generate motion sequences conditioned on many modalities. MotionBind achieves state-of-the-art or competitive performance across motion reconstruction, cross-modal retrieval, zero-shot action recognition, and text-to-motion generation benchmarks.

Knowledge-based visual question answering (KB-VQA) requires visual language models (VLMs) to integrate visual understanding with external knowledge retrieval. Although retrieval-augmented generation (RAG) achieves significant advances in this task by combining knowledge-base querying, it still struggles with the quality of multimodal queries and the relevance of retrieved results. To overcome these challenges, we propose a novel three-stage method, termed Wiki-PRF, including Processing, Retrieval and Filtering stages.

Contrastive Language-Image Pretraining (CLIP) achieves strong generalization in vision-language tasks by aligning images and texts in a shared embedding space. However, recent findings show that CLIP-like models still underutilize fine-grained semantic signals in text, and this issue becomes even more pronounced when dealing with long and detailed captions. This stems from CLIP's training objective, which optimizes only global image-text similarity and overlooks tokenlevel supervision--limiting its ability to achieve fine-grained visual-text alignment. To address this, we propose SuperCLIP, a simple yet effective framework that augments contrastive learning with classification-based supervision. By adding only a lightweight linear layer to the vision encoder, SuperCLIP leverages tokenlevel cues to enhance visual-textual alignment -- with just a 0.077% increase in total FLOPs, and no need for additional annotated data. Experiments show that SuperCLIP consistently improves zero-shot classification, image-text retrieval, and purely visual tasks. These gains hold regardless of whether the model is trained on original web data or rich re-captioned data, demonstrating SuperCLIP's ability to recover textual supervision in both cases. Furthermore, SuperCLIP alleviates CLIP's small-batch performance drop through classification-based supervision that avoids reliance on large batch sizes.

Prompt Generator Quality Filter 1 Image2 Image1 2 Person retrieval has attracted rising attention.

We introduce the Region Encoder Network (REN), a fast and effective model for generating region-based image representations using point prompts. Recent methods combine class-agnostic segmenters (e.g., SAM) with patch-based image encoders (e.g., DINO) to produce compact and effective region representations, but they suffer from high computational cost due to the segmentation step. REN bypasses this bottleneck using a lightweight module that directly generates region tokens, enabling 60 faster token generation with 35 less memory, while also improving token quality. It uses a few cross-attention blocks that take point prompts as queries and features from a patch-based image encoder as keys and values to produce region tokens that correspond to the prompted objects. We train REN with three popular encoders--DINO, DINOv2, and OpenCLIP--and show that it can be extended to other encoders without dedicated training. We evaluate REN on semantic segmentation and retrieval tasks, where it consistently outperforms the original encoders in both performance and compactness, and matches or exceeds SAMbased region methods while being significantly faster. Notably, REN achieves state-of-the-art results on the challenging Ego4DVQ2D benchmark and outperforms proprietary LMMs on Visual Haystacks' single-needle challenge. The code and pretrained models are available at https://github.com/savya08/ren.

Cross-modal hashing aims to efficiently retrieve information across different modalities by mapping data into compact hash codes. However, most existing methods assume access to fully accurate supervision, which rarely holds in real-world scenarios. In fact, annotations are often redundant, i.e., each sample is associated with a set of candidate labels that includes both ground-truth labels and redundant noisy labels. Treating all annotated labels as equally valid introduces two critical issues: (1) the sparse presence of true labels within the label set is not explicitly addressed, leading to overfitting on redundant noisy annotations; (2) redundant noisy labels induce spurious similarities that distort semantic alignment across modalities and degrade the quality of the hash space. To address these challenges, we propose that effective cross-modal hashing requires explicitly identifying and leveraging the true label subset within all candidate annotations.

Understanding and reasoning over long videos pose significant challenges for large video language models (LVLMs) due to the difficulty in processing intensive video tokens beyond context window and retaining long-term sequential information. Retrieval-Augmented Generation (RAG) has demonstrated effectiveness in processing long context for Large Language Models (LLMs); however, applying RAG to long video faces challenges such as disrupted temporal dependencies and inclusion of irrelevant information that can hinder accurate reasoning. To address these limitations, we propose Vgent, a novel graph-based retrieval-reasoning-augmented generation framework to enhance LVLMs for long video understanding. Our approach introduces two key innovations: (i) It represents videos by structured graphs with semantic relationships across video clips preserved to improve retrieval effectiveness.

Most multimodal models treat every negative pair alike, ignoring the ambiguous negatives that differ from the positive by only a small detail. We propose BoundaryA ware Curriculum with Local Attention(BACL), a lightweight add-on that turns these borderline cases into a curriculum signal. ABoundary-aware Negative Sampler gradually raises difficulty, while a Contrastive Local Attention loss highlights where the mismatch occurs. The two modules are fully differentiable and work with any off-the-shelf dual encoder. Theory predicts a fast O(1/n) error rate; practice shows up to +32 % R@1 over CLIP and new SOTA on four large-scale benchmarks, all without extra labels.

Text-3DScene Retrieval (T3SR) aims to retrieve relevant scenes using linguistic queries. Although traditional T3SR methods have made significant progress in capturing fine-grained associations, they implicitly assume that query descriptions are information-complete. In practical deployments, however, limited by the capabilities of users and models, it is difficult or even impossible to directly obtain a perfect textual query suiting the entire scene and model, thereby leading to performance degradation. To address this issue, we propose a novel Interactive Text-3D Scene Retrieval Method (IDeal), which promotes the enhancement of the alignment between texts and 3D scenes through continuous interaction. To achieve this, we present an Interactive Retrieval Refinement framework (IRR), which employs a questioner to pose contextually relevant questions to an answerer in successive rounds that either promote detailed probing or encourage exploratory divergence within scenes. Upon the iterative responses received from the answerer, IRR adopts a retriever to perform both feature-level and semantic-level information fusion, facilitating scene-level interaction and understanding for more precise re-rankings. To bridge the domain gap between queries and interactive texts, we propose an Interaction Adaptation Tuning strategy (IAT).