4DGS360 addresses the ill-posed challenge of reconstructing dynamic objects from monocular video by tackling a critical failure mode: existing methods rely on 2D-native priors that overfit to visible surfaces and cannot reconstruct occluded regions at extreme viewpoints (>90°). The authors propose AnchorTAP3D, a hybrid 3D tracker that leverages high-confidence 2D track points as spatial-temporal anchors to stabilize long-term tracking and resolve depth ambiguity in occluded areas. Combined with a new iPhone360 benchmark featuring test cameras up to 135° from training views, the method enables coherent 360° 4D reconstruction without diffusion priors.

Omni-WorldBench addresses the gap between passive video generation metrics and active world model evaluation by focusing on interactive response—how actions causally drive state transitions across space and time. It introduces Omni-WorldSuite, a 1,068-prompt hierarchical taxonomy spanning three interaction levels (single-object to global environmental effects), and Omni-Metrics, an agent-based evaluation protocol that aggregates Interaction Effect Fidelity, Generated Video Quality, and Camera-Object Controllability into an adaptive AgenticScore.

AdaEdit tackles the injection dilemma in flow-based image editing, where source feature injection preserves backgrounds but suppresses novel content generation. The authors propose two training-free adaptations: a Progressive Injection Schedule using continuous decay functions (sigmoid, cosine, linear) instead of binary cutoffs, and Channel-Selective Latent Perturbation that applies per-channel AdaIN based on distributional gaps between inverted and random latents. Extensive experiments on PIE-Bench show AdaEdit improves background preservation metrics by 8.7% LPIPS reduction versus ProEdit while maintaining competitive CLIP scores.

This paper presents PhotoBeamSolver, a hybrid system that converts hand-drawn beam diagrams into analytical structural solutions by combining computer vision with large language models. The core idea uses a custom-trained YOLO-based detector to identify supports and loads from images, feeding a symbolic solver that computes shear, moment, and deflection diagrams. While targeted at academic and quick professional verification tasks, the work highlights the challenges of integrating deep learning into safety-critical structural engineering workflows.

This paper challenges the long-held assumption that infrared and visible image fusion (IVIF) requires strictly paired training data. The authors propose UnPaired and Arbitrarily Paired Training Paradigms (UPTP and APTP), demonstrating that pixel-level self-supervision enables training on unaligned cross-modal combinations. By reformulating the maximum likelihood objective to treat infrared and visible images as independent variables, they show that a base dataset of $N$ pairs can be expanded to $N^2$ trainable combinations, potentially reducing collection costs while improving generalization.

Articulated object reconstruction typically requires either multi-view capture of discrete states or monocular video with a strict static-base-part assumption, limiting practical deployment. FreeArtGS introduces a "free-moving" setting where both joint angles and object poses vary arbitrarily during capture, using only a monocular RGB-D video. The method combines motion-based part segmentation via point tracking priors with joint estimation and 3D Gaussian Splatting optimization to jointly reconstruct geometry, appearance, and articulation.

Cardiac ultrasound view acquisition is notoriously operator-dependent, limiting reproducibility and access. This paper proposes an anatomical prior (AP)-driven framework that unifies cardiac structure segmentation with autonomous probe adjustment. The core innovation is a spatial-relation graph (SRG) module that injects spatial-topological constraints into YOLO-based segmentation, coupled with an RL formulation where states and rewards are built from quantifiable anatomical features drawn from Gaussian priors. The work matters because it offers an interpretable alternative to black-box end-to-end methods, potentially enabling zero-shot sim-to-real deployment for robotic echocardiography.

3D fragment reassembly becomes challenging at scale because incorrect contact adjacencies trigger cascading failures. This paper proposes SARe, a generative framework that explicitly models contact structure by jointly predicting fracture-surface tokens and inter-fragment adjacency graphs, paired with an inference-time refinement stage that anchors reliable substructures to correct uncertain regions. The work demonstrates state-of-the-art results across synthetic and real fracture datasets, with notable improvements in the many-fragment ($K$) regime.

This paper tackles unregistered hyperspectral-multispectral image fusion (HMF), where spatially misaligned images with partial overlap must be mutually super-resolved without training data or co-registration. The authors propose FRESCO, a two-stage unsupervised framework that uses coupled block-term tensor decomposition (BTD) for MSI spectral super-resolution and latent-space adversarial learning for HSI spatial super-resolution. The work is notable for offering the first theoretical recoverability guarantees in the unregistered setting, addressing a practically important gap in remote sensing.

Video facial expression recognition (FER) suffers from severe subject-specific distribution shifts that degrade CLIP model performance at test time. This paper proposes TTA-CaP, a gradient-free test-time adaptation method that personalizes models using three coordinated caches—a fixed source-domain prototype cache, a dynamic positive target cache for reliable samples, and a negative cache for uncertain predictions—coupled with a tri-gate filtering mechanism to prevent error accumulation.

DepthTCM tackles depth map compression by combining physics-inspired Multiwavelength Depth (MWD) encoding—mapping depth to sinusoidal 3-channel images—with global 4-bit quantization and a Transformer-CNN mixed learned codec. The core claim is that this hybrid approach reshapes depth statistics into a form amenable to modern learned image compression, achieving 60% bitrate reduction over prior MWD methods while maintaining >99% geometric accuracy.

The paper addresses the quadratic complexity of transformer attention and limited local detail extraction in RGB-D Salient Object Detection (SOD). It proposes STENet, which introduces superpixels as intermediate tokens to reduce computational overhead while preserving structural coherence. The core idea replaces global pixel-to-pixel attention with two modules: one for pixel-to-superpixel global enhancement and another for intra-superpixel local refinement, aiming to balance efficiency and accuracy.

BanglaVerse introduces a culturally grounded benchmark evaluating vision-language models on Bengali culture across standard Bangla, four historically linked languages, and five regional dialects. Built from 1,152 manually curated images expanded to ~32.3K artifacts, the work reveals that standard Bangla evaluation substantially overestimates model capabilities compared to dialectal settings. The core finding—that missing cultural knowledge, not visual grounding alone, is the primary bottleneck—challenges conventional multimodal evaluation practices for underrepresented languages.

Video subtitle removal traditionally requires expensive per-frame mask annotations and external detection modules during both training and inference. CLEAR introduces a two-stage mask-free framework that decouples prior extraction (via self-supervised disentangled feature learning) from generative refinement (via LoRA-adapted diffusion with adaptive weighting). The method claims to train only 0.77% of base model parameters while achieving +6.77dB PSNR gains and zero-shot generalization across six languages without ground-truth masks at inference.

daVinci-MagiHuman tackles joint audio-video generation using a refreshingly simple single-stream Transformer that processes text, video, and audio tokens through self-attention only---avoiding the cross-attention and fusion modules common in competing multi-stream architectures. The model achieves strong human-centric generation quality across six languages while delivering impressive inference speed: 2 seconds for a 5-second 256p video on an H100.

Beta-KD tackles the problem of balancing data supervision against teacher guidance when distilling multimodal large language models. The authors frame knowledge distillation as Bayesian MAP estimation with teacher-informed Gibbs priors over student activations, deriving a closed-form uncertainty-aware weighting mechanism via Laplace approximation. This eliminates manual tuning of loss weights and achieves consistent improvements across six VQA benchmarks.

This paper addresses hypertension screening from inexpensive retinal fundus images by distilling knowledge from high-fidelity brain MRI—without requiring paired acquisitions from the same patients. The proposed Clinical Graph-Mediated Distillation (CGMD) constructs a clinical similarity graph using shared biomarkers (age, labs, etc.) to bridge disjoint MRI and fundus cohorts, propagates MRI teacher embeddings over the graph to impute patient-specific targets for fundus patients, and trains a fundus student with supervised, prior, and relational distillation losses. The approach aims to capture subtle vascular signals in fundus images by leveraging MRI-derived markers of small-vessel disease.

HMS-VesselNet addresses the challenge of segmenting thin peripheral retinal vessels in fundus images—a critical task for early diabetic retinopathy detection where standard overlap losses fail due to class imbalance and topological fragmentation. The paper proposes a four-scale hierarchical Attention U-Net architecture with learned fusion weights, combining Dice, binary cross-entropy, and centerline Dice ($\text{clDice}$) losses alongside hard example mining to boost sensitivity on sub-2-pixel vessels. Evaluated on 68 images from DRIVE, STARE, and CHASE_DB1 via 5-fold cross-validation and leave-one-dataset-out protocols, the model achieves $90.78\pm1.42\%$ Sensitivity, demonstrating that explicit topology preservation and targeted hard example oversampling can recover fine vascular structures missed by standard area-based losses.

Vision-Language Models face escalating safety risks from adversarial jailbreak attacks that bypass alignment via manipulated visual inputs. This paper introduces NullSteer, a training-free defense that applies activation steering constrained to the null space of benign representations—mathematically guaranteeing that safe inputs remain unchanged while harmful activations are redirected toward refusal semantics. The approach aims to solve the over-refusal problem plaguing existing steering methods, offering a principled trade-off between robust safety and preserved utility.

Most visual counting benchmarks focus on rigid objects like crowds and vehicles, leaving fine-grained biological counting understudied. This paper introduces TPC–268, a dataset of 10,000 images spanning 268 countable plant categories across 242 species, annotated with full Linnaean taxonomies and biological organization levels. By framing plant counting as class-agnostic counting with taxonomic constraints, the authors provide a testbed for evaluating hierarchical generalization in vision models.