Cross-Modal Knowledge Distillation

Detecting amyloid-$β$ (A$β$) positivity is crucial for early diagnosis of Alzheimer's disease but typically requires PET imaging, which is costly, invasive, and not widely accessible, limiting its use for population-level screening. We address this gap by proposing a PET-guided knowledge distillation framework that enables A$β$ prediction from MRI alone, without requiring non-imaging clinical covariates or PET at inference. Our approach employs a BiomedCLIP-based teacher model that learns PET-MRI alignment via cross-modal attention and triplet contrastive learning with PET-informed (Centiloid-aware) online negative sampling. An MRI-only student then mimics the teacher via feature-level and logit-level distillation. Evaluated across four MRI contrasts (T1w, T2w, FLAIR, T2*) and two independent datasets, our approach demonstrates effective knowledge transfer (best AUC: 0.74 on OASIS-3, 0.68 on ADNI) while maintaining interpretability and eliminating the need for clinical variables. Saliency analysis confirms that predictions focus on anatomically relevant cortical regions, supporting the clinical viability of PET-free A$β$ screening. Code is available at https://github.com/FrancescoChiumento/pet-guided-mri-amyloid-detection.

Spatial transcriptomics provides a molecularly rich description of tissue organization, enabling unsupervised discovery of tissue niches -- spatially coherent regions of distinct cell-type composition and function that are relevant to both biological research and clinical interpretation. However, spatial transcriptomics remains costly and scarce, while H&E histology is abundant but carries a less granular signal. We propose to leverage paired spatial transcriptomics and H&E data to transfer transcriptomics-derived niche structure to a histology-only model via cross-modal distillation. Across multiple tissue types and disease contexts, the distilled model achieves substantially higher agreement with transcriptomics-derived niche structure than unsupervised morphology-based baselines trained on identical image features, and recovers biologically meaningful neighborhood composition as confirmed by cell-type analysis. The resulting framework leverages paired spatial transcriptomic and H&E data during training, and can then be applied to held-out tissue regions using histology alone, without any transcriptomic input at inference time.

Robust multimodal human sensing must overcome the critical challenge of missing modalities. Two principal barriers are the Representation Gap between heterogeneous data and the Contamination Effect from low-quality modalities. These barriers are causally linked, as the corruption introduced by contamination fundamentally impedes the reduction of representation disparities. In this paper, we propose PTA, a novel "Purify-then-Align" framework that solves this causal dependency through a synergistic integration of meta-learning and knowledge diffusion. To purify the knowledge source, PTA first employs a meta-learning-driven weighting mechanism that dynamically learns to down-weight the influence of noisy, low-contributing modalities. Subsequently, to align different modalities, PTA introduces a diffusion-based knowledge distillation paradigm in which an information-rich clean teacher, formed from this purified consensus, refines the features of each student modality. The ultimate payoff of this "Purify-then-Align" strategy is the creation of exceptionally powerful single-modality encoders imbued with cross-modal knowledge. Comprehensive experiments on the large-scale MM-Fi and XRF55 datasets, under pronounced Representation Gap and Contamination Effect, demonstrate that PTA achieves state-of-the-art performance and significantly improves the robustness of single-modality models in diverse missing-modality scenarios.

Very High Resolution (VHR) forest structure data at individual-tree scale is essential for carbon, biodiversity, and ecosystem monitoring. Still, airborne LiDAR remains costly and infrequent despite being the reference for forest structure metrics like Canopy Height Model (CHM), Plant Area Index (PAI), and Foliage Height Diversity (FHD). We propose FSKD: a LiDAR-to-RGB-Infrared (RGBI) knowledge distillation (KD) framework in which a multi-modal teacher fuses RGBI imagery with LiDAR-derived planar metrics and vertical profiles via cross-attention, and an RGBI-only SegFormer student learns to reproduce these outputs. Trained on 384 $km^2$ of forests in Saxony, Germany (20 cm ground sampling distance (GSD)) and evaluated on eight geographically distinct test tiles, the student achieves state-of-the-art (SOTA) zero-shot CHM performance (MedAE 4.17 m, $R^2$=0.51, IoU 0.87), outperforming HRCHM/DAC baselines by 29--46% in MAE (5.81 m vs. 8.14--10.84 m) with stronger correlation coefficients (0.713 vs. 0.166--0.652). Ablations show that multi-modal fusion improves performance by 10--26% over RGBI-only training, and that asymmetric distillation with appropriate model capacity is critical. The method jointly predicts CHM, PAI, and FHD, a multi-metric capability not provided by current monocular CHM estimators, although PAI/FHD transfer remains region-dependent and benefits from local calibration. The framework also remains effective under temporal mismatch (winter LiDAR, summer RGBI), removing strict co-acquisition constraints and enabling scalable 20 cm operational monitoring for workflows such as Digital Twin Germany and national Digital Orthophoto programs.

Turning a pretrained language model (LM) into a vision-language model (VLM) through multimodal fine-tuning often erodes its native language ability, a form of catastrophic forgetting that shows up even on text-only tasks. This loss is hard to undo with further fine-tuning, and existing remedies add adapters or alignment modules that increase architectural complexity and inference cost. We propose LinguDistill, an adapter-free knowledge distillation method that uses the original frozen LM as the teacher during multimodal post-training. To let a text-only teacher supervise vision-conditioned outputs, we introduce layer-wise KV-cache sharing, which exposes the teacher to the student's multimodal representations without changing either architecture. We then apply distillation selectively, on language-heavy data only, so the teacher restores linguistic ability while the student keeps its visual grounding on document and OCR tasks. LinguDistill recovers the language and knowledge performance lost during multimodal fine-tuning, matching the original VLM on average over text-only benchmarks (ARC, HellaSwag) and exceeding it on ScienceQA, while keeping vision-heavy performance close to standard fine-tuning. Since the teacher is dropped after training, the final model adds no parameters and no inference cost. More broadly, our results show that a model's own pre-adaptation backbone is a practical teacher for undoing forgetting, suggesting a simple recipe for keeping language ability intact as models are extended to new modalities.

We study a systems-level visual inference problem: using an expensive privileged modality during training while preserving a fixed-cost, single-modality deployment path. We present JDCNet, a confidence-gated CT-to-X-ray distillation framework in which the CT teacher supplies an auxiliary hard or temperature-scaled target only on training samples whose teacher confidence exceeds a threshold; at deployment the student takes X-ray input alone and matches the parameter, MAC, and latency profile of the supervised X-ray baseline. On a 510-patient same-patient paired BIMCV cohort with patient-level 5-fold cross-validation, two JDCNet configurations clear a fixed transfer gate against the supervised ResNet-18 baseline: 3-slice soft-KL supervision yields $Δ\mathrm{BA}{=}{+}0.035$ ($95\%$ CI $[{+}0.011,{+}0.057]$) and mid-slice hard supervision yields $+0.033$ ($[{+}0.007,{+}0.058]$). Under the same splits and gate, logit distillation, gated logit distillation, contrastive alignment, attention transfer, feature hints, BiomedCLIP fine-tuning, and a module-augmented variant do not pass. Confidence-gated auxiliary targets are therefore a more transferable channel than uniformly softened CT logits; the evidence is bounded to one paired cohort, so external paired-cohort replication is required before any deployment claim.

Non-contact automatic deception detection remains challenging because visual and auditory deception cues often lack stable cross-subject patterns. In contrast, galvanic skin response (GSR) provides more reliable physiological cues and has been widely used in contact-based deception detection. In this work, we leverage stable deception-related knowledge in GSR to guide representation learning in non-contact modalities through cross-modal knowledge distillation. A key obstacle, however, is the lack of a suitable dataset for this setting. To address this, we introduce MuDD, a large-scale Multimodal Deception Detection dataset containing recordings from 130 participants over 690 minutes. In addition to video, audio, and GSR, MuDD also provides Photoplethysmography, heart rate, and personality traits, supporting broader scientific studies of deception. Based on this dataset, we propose GSR-guided Progressive Distillation (GPD), a cross-modal distillation framework for mitigating the negative transfer caused by the large modality mismatch between GSR and non-contact signals. The core innovation of GPD is the integration of progressive feature-level and digit-level distillation with dynamic routing, which allows the model to adaptively determine how teacher knowledge should be transferred during training, leading to more stable cross-modal knowledge transfer. Extensive experiments and visualizations show that GPD outperforms existing methods and achieves state-of-the-art performance on both deception detection and concealed-digit identification.

Contrastive Language-Image Pre-training (CLIP) demonstrates strong zero-shot generalization, but due to substantial computational and memory costs, distillation into lightweight models is required. Existing relational objectives do not explicitly model multidirectional relationships between teacher and student embeddings, potentially leaving the geometric relationships insufficiently constrained. This may disrupt the modality-gap structure important for zero-shot transfer. To address these limitations, we propose a relational distillation framework, CLIP-RD, which introduces two relational methods, Vertical Relational Distillation (VRD) and Cross Relational Distillation (XRD). VRD aligns teacher-student intra-modal similarity distributions to enforce consistent distillation strength across image and text embeddings. Meanwhile, XRD aligns the teacher-image-student-text and teacher-text-student-image similarity distributions to impose bidirectional cross-modal symmetry. By jointly modeling these multidirectional relational structures, CLIP-RD aligns the student's embedding geometry more faithfully to the teacher's, outperforming CLIP-KD by 1.8%p. This performance improvement is maintained across diverse architectures, teacher scales, retrieval tasks, downstream tasks, and corruption settings, with negligible additional training-time overhead.

Knowledge distillation transfers knowledge from large teacher models to smaller students for efficient inference. While existing methods primarily focus on distillation strategies, they often overlook the importance of enhancing teacher knowledge quality. In this paper, we propose Text-guided Multi-view Knowledge Distillation (TMKD), which leverages dual-modality teachers, a visual teacher and a text teacher (CLIP), to provide richer supervisory signals. Specifically, we enhance the visual teacher with multi-view inputs incorporating visual priors (edge and high-frequency features), while the text teacher generates semantic weights through prior-aware prompts to guide adaptive feature fusion. Additionally, we introduce vision-language contrastive regularization to strengthen semantic knowledge in the student model. Extensive experiments on five benchmarks demonstrate that TMKD consistently improves knowledge distillation performance by up to 4.49%, validating the effectiveness of our dual-teacher multi-view enhancement strategy. Code is available at https://anonymous.4open.science/r/TMKD-main-44D1.

Unstructured Electronic Health Record (EHR) data, such as clinical notes, contain clinical contextual observations that are not directly reflected in structured data fields. This additional information can substantially improve model learning. However, due to their unstructured nature, these data are often unavailable or impractical to use when deploying a model. We introduce a multimodal learning framework that leverages unstructured EHR data during training while producing a model that can be deployed using only structured EHR data. Using a cohort of 3,466 children evaluated for late talking, we generated note embeddings with BioClinicalBERT and encoded structured embeddings from demographics and medical codes. A note-based teacher model and a structured-only student model were jointly trained using contrastive learning and contrastive knowledge distillation loss, producing a strong classifier (AUROC = 0.985). Our proposed model reached an AUROC of 0.705, outperforming the structured-only baseline of 0.656. These results demonstrate that incorporating unstructured data during training enhances the model's capacity to identify task-relevant information within structured EHR data, enabling a deployable structured-only phenotype model.

Retinal fundus imaging enables low-cost and scalable hypertension (HTN) screening, but HTN-related retinal cues are subtle, yielding high-variance predictions. Brain MRI provides stronger vascular and small-vessel-disease markers of HTN, yet it is expensive and rarely acquired alongside fundus images, resulting in modality-siloed datasets with disjoint MRI and fundus cohorts. We study this unpaired MRI-fundus regime and introduce Clinical Graph-Mediated Distillation (CGMD), a framework that transfers MRI-derived HTN knowledge to a fundus model without paired multimodal data. CGMD leverages shared structured biomarkers as a bridge by constructing a clinical similarity kNN graph spanning both cohorts. We train an MRI teacher, propagate its representations over the graph, and impute brain-informed representation targets for fundus patients. A fundus student is then trained with a joint objective combining HTN supervision, target distillation, and relational distillation. Experiments on our newly collected unpaired MRI-fundus-biomarker dataset show that CGMD consistently improves fundus-based HTN prediction over standard distillation and non-graph imputation baselines, with ablations confirming the importance of clinically grounded graph connectivity. Code is available at https://github.com/DillanImans/CGMD-unpaired-distillation.

Recent vision-language models such as CLIP provide strong cross-modal alignment, but current CLIP-guided ReID pipelines rely on global features and fixed prompts. This limits their ability to capture fine-grained attribute cues and adapt to diverse appearances. We propose ALADIN, an attribute-language distillation network that distills knowledge from a frozen CLIP teacher to a lightweight ReID student. ALADIN introduces fine-grained attribute-local alignment to establish adaptive text-visual correspondence and robust representation learning. A Scene-Aware Prompt Generator produces image-specific soft prompts to facilitate adaptive alignment. Attribute-local distillation enforces consistency between textual attributes and local visual features, significantly enhancing robustness under occlusions. Furthermore, we employ cross-modal contrastive and relation distillation to preserve the inherent structural relationships among attributes. To provide precise supervision, we leverage Multimodal LLMs to generate structured attribute descriptions, which are then converted into localized attention maps via CLIP. At inference, only the student is used. Experiments on Market-1501, DukeMTMC-reID, and MSMT17 show improvements over CNN-, Transformer-, and CLIP-based methods, with better generalization and interpretability.

Vision-based policies are widely applied in robotics for tasks such as manipulation and locomotion. On lightweight mobile robots, however, they face a trilemma of limited scene transferability, restricted onboard computation resources, and sensor hardware cost. To address these issues, we propose a knowledge distillation approach that transfers knowledge from an information-rich, appearance invariant omniview depth policy to a lightweight monocular policy. The key idea is to train the student not only to mimic the expert actions but also to align with the latent embeddings of the omni view depth teacher. Experiments demonstrate that omni-view and depth inputs improve the scene transfer and navigation performance, and that the proposed distillation method enhances the performance of a singleview monocular policy, compared with policies solely imitating actions. Real world experiments further validate the effectiveness and practicality of our approach. Code will be released publicly.

Recent adapter-based CLIP tuning (e.g., Tip-Adapter) is a strong few-shot learner, achieving efficiency by caching support features for fast prototype matching. However, these methods rely on global uni-modal feature vectors, overlooking fine-grained patch relations and their structural alignment with class text. To bridge this gap without incurring inference costs, we introduce a novel asymmetric training-only framework. Instead of altering the lightweight adapter, we construct a high-capacity auxiliary Heterogeneous Graph Teacher that operates solely during training. This teacher (i) integrates multi-scale visual patches and text prompts into a unified graph, (ii) performs deep cross-modal reasoning via a Modality-aware Graph Transformer (MGT), and (iii) applies discriminative node filtering to extract high-fidelity class features. Crucially, we employ a cache-aware dual-objective strategy to supervise this relational knowledge directly into the Tip-Adapter's key-value cache, effectively upgrading the prototypes while the graph teacher is discarded at test time. Thus, inference remains identical to Tip-Adapter with zero extra latency or memory. Across standard 1-16-shot benchmarks, our method consistently establishes a new state-of-the-art. Ablations confirm that the auxiliary graph supervision, text-guided reasoning, and node filtering are the essential ingredients for robust few-shot adaptation. Code is available at https://github.com/MR-Sherif/TOGA.git.

Adapting vision-language models to remote sensing imagery remains challenging due to two key factors: limited semantic coverage in textual representations and insufficient adaptability of visual features. These issues are particularly significant in aerial scenes, which involve various visual appearances and fine-grained object distinctions. We propose AVION, a knowledge distillation framework tailored for remote sensing adaptation of vision-language models. The teacher module constructs semantically rich textual prototypes by collecting descriptions from a large language model and verifying validity using remote sensing image features. The student module integrates lightweight and learnable prompts into both vision and language encoders, guided by the teacher to align embeddings and their cross-modal relationships. Once trained, the student operates independently during inference. Experiments on six optical remote sensing benchmarks show that AVION improves few-shot classification and base-class accuracy without degrading generalization to novel categories. It also enhances mean recall for cross-modal retrieval, with minimal additional trainable parameters.

Knowledge distillation (KD) methods are pivotal in compressing large pre-trained language models into smaller models, ensuring computational efficiency without significantly dropping performance. Traditional KD techniques assume homogeneity in modalities between the teacher (source) and the student (target) models. On the other hand, existing multimodal knowledge distillation methods require modality-specific pre-training of the teacher model, which is computationally infeasible in most cases. In this paper, we introduce ARMADA, an efficient cross-modal knowledge distillation framework designed to transfer knowledge from large vision-language models, including black-box models, to language-only models. Unlike existing KD techniques that rely on the internal structures of multimodal teachers or require computationally expensive pre-training, ARMADA leverages novel alignment techniques to distil knowledge without altering the teacher model, ensuring efficiency and scalability. We empirically validate ARMADA on twelve natural language understanding, eight complex generative reasoning and five instruction-tuning tasks, demonstrating consistent performance improvements in large models such as DeBERTa-v2-1.4B, OPT-1.3B, LLaMA-{3B, 7B, 8B}. ARMADA achieves up to 3.4% improvement on language understanding tasks and 2.6% boost in generative reasoning, all without requiring expensive multimodal pre-training or fine-tuning of the teacher model. Our findings challenge conventional knowledge distillation paradigms by demonstrating that even vision-language models, despite lacking direct textual understanding, can significantly enhance language models when distilled appropriately.

Multimodal fake news detection is crucial for mitigating societal disinformation. Existing approaches attempt to address this by fusing multimodal features or leveraging Large Language Models (LLMs) for advanced reasoning. However, these methods suffer from serious limitations, including a lack of comprehensive multi-view judgment and fusion, and prohibitive reasoning inefficiency due to the high computational costs of LLMs. To address these issues, we propose LLM-Guided Multi-View Reasoning Distillation for Fake News Detection ( LLM-MRD), a novel teacher-student framework. The Student Multi-view Reasoning module first constructs a comprehensive foundation from textual, visual, and cross-modal perspectives. Then, the Teacher Multi-view Reasoning module generates deep reasoning chains as rich supervision signals. Our core Calibration Distillation mechanism efficiently distills this complex reasoning-derived knowledge into the efficient student model. Experiments show LLM-MRD significantly outperforms state-of-the-art baselines. Notably, it demonstrates a comprehensive average improvement of 5.19% in ACC and 6.33% in F1-Fake when evaluated across all competing methods and datasets. Our code is available at https://github.com/Nasuro55/LLM-MRD

Multi-view 3D detection with bird's eye view (BEV) is crucial for autonomous driving and robotics, but its robustness in real-world is limited as it struggles to predict accurate depth values. A mainstream solution, cross-modal distillation, transfers depth information from LiDAR to camera models but also unintentionally transfers depth-irrelevant information (e.g. LiDAR density). To mitigate this issue, we propose RayD3D, which transfers crucial depth knowledge along the ray: a line projecting from the camera to true location of an object. It is based on the fundamental imaging principle that predicted location of this object can only vary along this ray, which is finally determined by predicted depth value. Therefore, distilling along the ray enables more effective depth information transfer. More specifically, we design two ray-based distillation modules. Ray-based Contrastive Distillation (RCD) incorporates contrastive learning into distillation by sampling along the ray to learn how LiDAR accurately locates objects. Ray-based Weighted Distillation (RWD) adaptively adjusts distillation weight based on the ray to minimize the interference of depth-irrelevant information in LiDAR. For validation, we widely apply RayD3D into three representative types of BEV-based models, including BEVDet, BEVDepth4D, and BEVFormer. Our method is trained on clean NuScenes, and tested on both clean NuScenes and RoboBEV with a variety types of data corruptions. Our method significantly improves the robustness of all the three base models in all scenarios without increasing inference costs, and achieves the best when compared to recently released multi-view and distillation models.

Monocular 3D object detection is a promising yet ill-posed task for autonomous vehicles due to the lack of accurate depth information. Cross-modality knowledge distillation could effectively transfer depth information from LiDAR to image-based network. However, modality gap between image and LiDAR seriously limits its accuracy. In this paper, we systematically investigate the negative transfer problem induced by modality gap in cross-modality distillation for the first time, including not only the architecture inconsistency issue but more importantly the feature overfitting issue. We propose a selective learning approach named MonoSTL to overcome these issues, which encourages positive transfer of depth information from LiDAR while alleviates the negative transfer on image-based network. On the one hand, we utilize similar architectures to ensure spatial alignment of features between image-based and LiDAR-based networks. On the other hand, we develop two novel distillation modules, namely Depth-Aware Selective Feature Distillation (DASFD) and Depth-Aware Selective Relation Distillation (DASRD), which selectively learn positive features and relationships of objects by integrating depth uncertainty into feature and relation distillations, respectively. Our approach can be seamlessly integrated into various CNN-based and DETR-based models, where we take three recent models on KITTI and a recent model on NuScenes for validation. Extensive experiments show that our approach considerably improves the accuracy of the base models and thereby achieves the best accuracy compared with all recently released SOTA models.

While the shift from cascaded dialogue systems to end-to-end (E2E) speech Large Language Models (LLMs) improves latency and paralinguistic modeling, E2E models often exhibit a significant performance degradation compared to their text-based counterparts. The standard Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) training methods fail to close this gap. To address this, we propose X-OPD, a novel Cross-Modal On-Policy Distillation framework designed to systematically align the capabilities of Speech LLMs to their text-based counterparts. X-OPD enables the Speech LLM to explore its own distribution via on-policy rollouts, where a text-based teacher model evaluates these trajectories and provides token-level feedback, effectively distilling teacher's capabilities into student's multi-modal representations. Extensive experiments across multiple benchmarks demonstrate that X-OPD significantly narrows the gap in complex tasks while preserving the model's inherent capabilities.

Pretraining for electroencephalogram (EEG) foundation models has predominantly relied on self-supervised masked reconstruction, a paradigm largely adapted from and inspired by the success of vision and language foundation models. However, unlike images and text, EEG datasets are notoriously expensive to collect and characterized by low signal-to-noise ratio. These challenges introduce difficulties in scaling the EEG foundation models and capturing the underlying neural semantics through reconstruction. In this work, we ask the question: can we stand on the shoulders of well-established foundation models from well-represented modalities to bootstrap the pretraining of EEG foundation models? We first demonstrate that mainstream foundation models, such as those from vision and time series, transfer surprisingly well to EEG domain. To this end, we propose the Multi-Teacher Distillation Pretraining (MTDP) framework for pretraining EEG foundation models via a two-stage multi-teacher distillation. In the first stage, we introduce a learnable gating network to fuse representations from diverse teachers (e.g., DINOv3 and Chronos) via a masked latent denoising objective. In the second stage, we distill the fused representation into an EEG foundation model. Extensive evaluations across 9 downstream tasks and 12 datasets demonstrate that our MTDP-based EEG foundation model outperforms its self-supervised counterparts while requiring only 25% of the pretraining data.

Reliable Alzheimer's disease (AD) diagnosis increasingly relies on multimodal assessments combining structural Magnetic Resonance Imaging (MRI) and Electronic Health Records (EHR). However, deploying these models is bottlenecked by modality missingness, as MRI scans are expensive and frequently unavailable in many patient cohorts. Furthermore, synthesizing de novo 3D anatomical scans from sparse, high-dimensional tabular records is technically challenging and poses severe clinical risks. To address this, we introduce MIRAGE, a novel framework that reframes the missing-MRI problem as an anatomy-guided cross-modal latent distillation task. First, MIRAGE leverages a Biomedical Knowledge Graph (KG) and Graph Attention Networks to map heterogeneous EHR variables into a unified embedding space that can be propagated from cohorts with real MRIs to cohorts without them. To bridge the semantic gap and enforce physical spatial awareness, we employ a frozen pre-trained 3D U-Net decoder strictly as an auxiliary regularization engine. Supported by a novel cohort-aggregated skip feature compensation strategy, this decoder acts as a rigorous structural penalty, forcing 1D latent representations to encode biologically plausible, macro-level pathological semantics. By exclusively utilizing this distilled "diagnostic-surrogate" representation during inference, MIRAGE completely bypasses computationally expensive 3D voxel reconstruction. Experiments demonstrate that our framework successfully bridges the missing-modality gap, improving the AD classification rate by 13% compared to unimodal baselines in cohorts without real MRIs.

Alzheimer's disease is a progressive neurodegenerative disorder in which mild cognitive impairment (MCI) precedes dementia. Structural MRI provides biomarkers but requires costly infrastructure, limiting population-scale deployment. Speech offers a non-invasive alternative, yet speech-only classifiers are developed independently of neuroimaging and lack biological grounding for CN-versus-MCI classification. We propose MINT (Multimodal Imaging-to-Speech Knowledge Transfer), a three-stage framework that transfers MRI-derived biomarker structure to speech during training. An MRI teacher defines a compact embedding space for CN-versus-MCI classification, while a residual projection head aligns speech representations to this space using a combined geometric loss. The frozen MRI classifier enables imaging-free inference. On ADNI-4, aligned speech achieves performance comparable to speech baselines, while multimodal fusion improves over MRI alone. Ablations identify dropout regularization and self-supervised pretraining as important design choices. To our knowledge, MINT is the first demonstration of MRI-to-speech knowledge transfer for early Alzheimer's screening without imaging at inference.

Multimodal learning that integrates genomics and histopathology has shown strong potential in cancer diagnosis, yet its clinical translation is hindered by the limited availability of paired histology-genomics data. Knowledge distillation (KD) offers a practical solution by transferring genomic supervision into histopathology models, enabling accurate inference using histology alone. However, existing KD methods rely on batch-local alignment, which introduces instability due to limited within-batch comparisons and ultimately degrades performance. To address these limitations, we propose Momentum Memory Knowledge Distillation (MoMKD), a cross-modal distillation framework driven by a momentum-updated memory. This memory aggregates genomic and histopathology information across batches, effectively enlarging the supervisory context available to each mini-batch. Furthermore, we decouple the gradients of the genomics and histology branches, preventing genomic signals from dominating histology feature learning during training and eliminating the modality-gap issue at inference time. Extensive experiments on the TCGA-BRCA benchmark (HER2, PR, and ODX classification tasks) and an independent in-house testing dataset demonstrate that MoMKD consistently outperforms state-of-the-art MIL and multimodal KD baselines, delivering strong performance and generalization under histology-only inference. Overall, MoMKD establishes a robust and generalizable knowledge distillation paradigm for computational pathology.

Foundation models are transforming Earth Observation (EO), yet the diversity of EO sensors and modalities makes a single universal model unrealistic. Multiple specialized EO foundation models (EOFMs) will likely coexist, making efficient knowledge transfer across modalities essential. Most existing EO pretraining relies on masked image modeling, which emphasizes local reconstruction but provides limited control over global semantic structure. To address this, we propose a dual-teacher contrastive distillation framework for multispectral imagery that aligns the student's pretraining objective with the contrastive self-distillation paradigm of modern optical vision foundation models (VFMs). Our approach combines a multispectral teacher with an optical VFM teacher, enabling coherent cross-modal representation learning. Experiments across diverse optical and multispectral benchmarks show that our model adapts to multispectral data without compromising performance on optical-only inputs, achieving state-of-the-art results in both settings, with an average improvement of 3.64 percentage points in semantic segmentation, 1.2 in change detection, and 1.31 in classification tasks. This demonstrates that contrastive distillation provides a principled and efficient approach to scalable representation learning across heterogeneous EO data sources. Project page: \textcolor{magenta}{https://wolfilip.github.io/DEO/}.

Purpose: Non-invasive grading of prostate cancer (PCa) from micro-ultrasound (micro-US) could expedite triage and guide biopsies toward the most aggressive regions, yet current models struggle to infer tissue micro-structure at coarse imaging resolutions. Methods: We introduce an unpaired histopathology knowledge-distillation strategy that trains a micro-US encoder to emulate the embedding distribution of a pretrained histopathology foundation model, conditioned on International Society of Urological Pathology (ISUP) grades. Training requires no patient-level pairing or image registration, and histopathology inputs are not used at inference. Results: Compared to the current state of the art, our approach increases sensitivity to clinically significant PCa (csPCa) at 60% specificity by 3.5% and improves overall sensitivity at 60% specificity by 1.2%. Conclusion: By enabling earlier and more dependable cancer risk stratification solely from imaging, our method advances clinical feasibility. Source code will be publicly released upon publication.

Vision-Language Models (VLMs) achieve strong multimodal performance but are costly to deploy, and post-training quantization often causes significant accuracy loss. Despite its potential, quantization-aware training for VLMs remains underexplored. We propose GRACE, a framework unifying knowledge distillation and QAT under the Information Bottleneck principle: quantization constrains information capacity while distillation guides what to preserve within this budget. Treating the teacher as a proxy for task-relevant information, we introduce confidence-gated decoupled distillation to filter unreliable supervision, relational centered kernel alignment to transfer visual token structures, and an adaptive controller via Lagrangian relaxation to balance fidelity against capacity constraints. Across extensive benchmarks on LLaVA and Qwen families, our INT4 models consistently outperform FP16 baselines (e.g., LLaVA-1.5-7B: 70.1 vs. 66.8 on SQA; Qwen2-VL-2B: 76.9 vs. 72.6 on MMBench), nearly matching teacher performance. Using real INT4 kernel, we achieve 3$\times$ throughput with 54% memory reduction. This principled framework significantly outperforms existing quantization methods, making GRACE a compelling solution for resource-constrained deployment. Code and data are available at: https://github.com/ForeverBlue816/GRACE.

Personalized object detection aims to adapt a general-purpose detector to recognize user-specific instances from only a few examples. Lightweight models often struggle in this setting due to their weak semantic priors, while large vision-language models (VLMs) offer strong object-level understanding but are too computationally demanding for real-time or on-device applications. We introduce MOCHA (Multi-modal Objects-aware Cross-arcHitecture Alignment), a distillation framework that transfers multimodal region-level knowledge from a frozen VLM teacher into a lightweight vision-only detector. MOCHA extracts fused visual and textual teacher's embeddings and uses them to guide student training through a dual-objective loss that enforces accurate local alignment and global relational consistency across regions. This process enables efficient transfer of semantics without the need for teacher modifications or textual input at inference. MOCHA consistently outperforms prior baselines across four personalized detection benchmarks under strict few-shot regimes, yielding a +10.1 average improvement, with minimal inference cost.

Recent advancements in vision-language models (VLMs) have leveraged large language models (LLMs) to achieve performance on par with closed-source systems like GPT-4V. However, deploying these models in real-world scenarios, particularly on resource-constrained devices, remains challenging due to their substantial computational demands. This has spurred interest in distilling knowledge from large VLMs into smaller, more efficient counterparts. A key challenge arises here from the diversity of VLM architectures, which are built on different LLMs and employ varying token types-differing in vocabulary size, token splits, and token index ordering. To address this challenge of limitation to a specific VLM type, we present Generation after Recalibration (GenRecal), a general-purpose distillation framework for VLMs. GenRecal incorporates a Recalibrator that aligns and adapts feature representations between heterogeneous VLMs, enabling effective knowledge transfer across different types of VLMs. Through extensive experiments on multiple challenging benchmarks, we demonstrate that GenRecal significantly improves baseline performances, eventually outperforming large-scale open- and closed-source VLMs.

Weakly supervised object localization (WSOL) models can predict both the object class and the spatial regions corresponding to the object, without requiring explicit bounding-box annotations. Given their reliance on classification objectives, traditional WSOL methods, like class activation mapping, tend to focus on the most discriminative object regions, often missing the full spatial extent. Although vision-language models like CLIP encode rich semantic priors, their global text and class-token embeddings are not explicitly aligned with local patch embeddings, limiting patch-level localization. Recent methods such as GenPrompt address this limitation, but at the cost of increased complexity, as they rely on conditional denoising and elaborate prompt-learning strategies. In this paper, we propose Text Distillation for Localization (TeD-Loc), which distills knowledge from CLIP text embeddings to patch embeddings through contrastive alignment, thereby enabling patch-level foreground/background localization. A localization-guided classification module is also introduced, which uses localization scores to aggregate foreground patch embeddings for joint classification and localization within a single model. In addition, a QR-based orthogonalization of class text embeddings is applied before distillation to improve discrimination for semantically similar classes. Extensive experiments show that TeD-Loc improves Top-1 Loc by ~5% on CUB and ILSVRC, and PxAP by ~31% on histopathology benchmarks, while achieving more efficient inference than GenPrompt.