Medical Vision-Language Models (Med-VLMs) for ultrasound analysis are vulnerable to subtle prompt variations that mimic real clinical communication patterns. This paper proposes a black-box attack framework using an LLM to generate minimal, clinically plausible text edits guided by Monte Carlo Tree Search (MCTS), requiring no access to the target model's weights or gradients. The study reveals that small adversarial rewrites can drastically degrade diagnostic QA accuracy—raising critical safety concerns for deploying such systems in point-of-care settings where prompt variability is inherent.

The paper addresses a fundamental limitation in 3D generation: image-conditioned models suffer from viewpoint bias and hallucinate unobserved regions, while text-conditioned models lack precise visual fidelity. The authors propose Text–Image Conditioned 3D Generation, a task requiring joint reasoning over visual exemplars and textual descriptions, and introduce TIGON—a minimalist dual-branch baseline that fuses separate image- and text-conditioned DiT backbones via zero-initialized cross-modal bridges and simple prediction averaging. This matters because it offers users more flexible control by combining pixel-aligned appearance cues with high-level semantic guidance.

This paper addresses video reasoning segmentation—segmenting objects in videos based on complex human instructions—by proposing TrajSeg, a unified framework built on Multimodal Large Language Models (MLLMs). The core innovation is bidirectional text-trajectory alignment, where the model learns both text-to-trajectory grounding and trajectory-to-text captioning, alongside a Frame-level Content Integration (FCI) module and a unified mask decoder that eliminates the need for separate key-frame and tracking models. The work matters because it simplifies training pipelines and aims to improve trajectory perception in dynamic video contexts.

3D-Layout-R1 tackles language-guided 3D spatial editing by training LLMs/VLMs to perform structured reasoning over explicit scene graphs. Instead of free-form chains-of-thought, the model outputs JSON graph edits that iteratively transform object poses and relations, combined with GRPO-based RL using dense 3D IoU and collision-aware rewards. This approach yields measurable gains in layout accuracy while maintaining interpretability across sorting, spatial alignment, and room-editing tasks.

Existing safety-critical scenario generation methods force collisions through brute-force perturbations, destroying trajectory realism. CounterScene reframes this as a counterfactual inference problem within diffusion-based BEV world models: given a safe scene, identify the single agent whose behavioral change would maximally increase collision risk, then minimally intervene on that agent alone via structured diffusion guidance. This targets the realism-adversarial trade-off by allowing danger to emerge through natural interaction propagation rather than global trajectory distortion.

Few-shot medical image segmentation (FSMIS) aims to segment anatomical structures with minimal annotations, but Segment Anything Model (SAM) based approaches suffer from over-segmentation due to ambiguous medical boundaries. This paper reformulates SAM-based FSMIS as a background-centric prompt localization task, proposing FoB (Focus on Background) to generate precise background prompts that constrain SAM’s predictions. By modeling contextual dependencies and ring-like structural priors, the method achieves state-of-the-art performance across CT, MRI, and dermatoscopic imaging while maintaining strong cross-domain generalization.

Multimodal skin cancer diagnosis with vision-language models faces a trilemma of computational cost, data scarcity, and black-box opacity. SkinCLIP-VL tackles this via a "frozen perception, adaptive reasoning" architecture that keeps CLIP frozen, adapts a quantized Qwen2.5-VL via LoRA, and introduces the Consistency-aware Focal Alignment (CFA) Loss to jointly handle class imbalance, cross-modal alignment, and calibration. The paper matters because it couples strong empirical performance with a clinician validation study, aiming to bridge the gap between AI accuracy and clinical trust.

GIDE addresses a key challenge in image editing: applying training-free editing techniques to Diffusion Large Language Models (DLLMs). Unlike continuous diffusion models where DDIM inversion is well-established, DLLMs use discrete tokenization that prevents direct application of standard noise inversion. GIDE introduces a three-stage framework (grounding, inversion, refinement) that enables precise localized editing via points, boxes, or text prompts while preserving background content. The significance lies in bridging discrete token spaces with high-fidelity inversion without additional training.

DGRNet addresses two critical gaps in brain tumor segmentation: reliable uncertainty quantification and under-utilization of radiology reports. The core idea transforms prediction disagreement among multiple lightweight view-specific adapters into an active signal that guides targeted refinement in ambiguous regions, integrated with clinical text conditioning. This approach achieves state-of-the-art accuracy on the TextBraTS benchmark while providing clinically meaningful uncertainty estimates calibrated to actual errors.

The paper presents a training-free pipeline for reconstructing instance-aware 3D scenes from 10-20 unposed RGB images and rendering novel views using diffusion. It combines MV-DUSt3R for geometry, SAM for 2D segmentation with warping-based cross-view unification, and the See3D diffusion model for inpainting holes in point-cloud projections. The system enables object-level editing by manipulating the point cloud directly, avoiding per-scene optimization.

OrbitStream addresses adaptive 360° video streaming for teleoperation by proposing a training-free framework that combines semantic scene understanding with robust control theory. It formulates viewport prediction as a Gravitational Viewport Prediction (GVP) problem where semantic objects (pedestrians, vehicles) generate potential fields that "attract" user gaze with task-relevant mass, while a Saturation-Based Proportional-Derivative (PD) Controller handles bitrate adaptation. This offers an interpretable, zero-shot alternative to black-box Deep Reinforcement Learning methods for safety-critical systems where deployment constraints prohibit lengthy training.

Medical vision-language models (VLMs) are increasingly evaluated for consistency—the invariance of predictions under paraphrased prompts—as a proxy for clinical reliability. This paper demonstrates that consistency alone is a fundamentally flawed safety metric because models can achieve perfect consistency by learning text shortcuts while completely ignoring the input image. The authors introduce a four-quadrant per-sample taxonomy that jointly evaluates consistency and image reliance, revealing that models optimized for low flip rates often shift samples into a 'Dangerous' quadrant where predictions are stable, accurate, and confident yet unchanged when the image is removed. Their findings expose a critical deployment trap: standard evaluation pipelines risk preferentially selecting models that appear reliable while being decision-invariant to visual evidence.

This paper investigates how vision-language models (VLMs) perform spatial reasoning—the binding of objects to spatial relations. It reveals that VLMs rely on two concurrent mechanisms: a dominant one where the vision encoder encodes object layout globally across visual tokens (extending into background regions), and a secondary one where the language model backbone forms ordering representations over object tokens. The finding that enhancing these vision-derived spatial representations improves performance without fine-tuning challenges the prevailing focus on LM backbones and highlights the critical role of vision encoders in multimodal reasoning.

This paper introduces BHDD, the first public benchmark dataset for handwritten Burmese digits. Myanmar script's distinctive circular letterforms—originally developed for writing on palm leaves—create recognition challenges distinct from Latin digits, with pairs like 0 and 1 differing only by whether a circle is closed. The authors release 87,561 verified images (28×28 grayscale, MNIST-compatible format) from over 150 contributors, with writer-independent train/test splits and baseline models reaching up to 99.83% accuracy.

HamVision proposes using damped harmonic oscillator dynamics as a structured inductive bias for medical image analysis. The core idea is that phase-space decomposition yields three representations—position $q$ (features), momentum $p$ (gradients), and energy $H = rac{1}{2}|z|^2$ (saliency)—that serve both segmentation and classification tasks without modifying the shared bottleneck. This physics-constrained approach aims to replace generic learned transformations with interpretable, dynamics-based feature extraction across diverse medical imaging modalities.

This paper addresses a subtle but critical issue in latent diffusion models (LDMs): VAE tokenizers tend to collapse latent variance toward zero to minimize reconstruction error, creating overly compact manifolds that are brittle against sampling perturbations. The authors propose a Variance Expansion (VE) loss that adaptively counteracts this collapse via an inverse-variance term $\mathcal{L}_{\text{var}} = 1/(\sigma^2 + \delta)$, allowing the latent space to absorb stochastic diffusion noise while maintaining reconstruction fidelity. The work achieves state-of-the-art FID 1.18 on ImageNet 256$\times$256 and provides both theoretical grounding and empirical validation across multiple architectures.

This paper tackles the visual perception gap in automated text layout generation. While existing Multimodal Large Language Models (MLLMs) generate layout code (SVG/JSON) to render text on images, they operate blind to the actual rendered output, producing layouts with overlapping text, poor contrast, or misalignment. The authors propose Visual Feedback Layout Model (VFLM), which closes the loop by rendering generated SVGs and feeding the visual results back to the model for iterative reflection and refinement. The framework uses a two-stage pipeline—cold-start supervised fine-tuning followed by reinforcement learning with GRPO—and introduces a specialized layout reward model trained on fine-grained quality hierarchies. A surprising finding is that simple outcome-based rewards outperform complex process-oriented rewards that explicitly encode step-wise incentives.

Brain tumor segmentation from MRI scans faces challenges because the three target sub-regions—Whole Tumor (WT), Tumor Core (TC), and Enhancing Tumor (ET)—have ambiguous visual boundaries. This paper proposes TextCSP, a hierarchical framework that integrates radiological reports by replacing the standard single global text embedding with sub-region-aware prompts and a soft cascade decoder that enforces the anatomical hierarchy $ET \subset TC \subset WT$. The method builds on the TextBraTS baseline and achieves modest gains on its paired MRI-text dataset.

dynActivation addresses the rigidity of fixed activation functions by introducing per-layer trainable scalars that interpolate between a base nonlinearity and a linear path. The method adds only two parameters per layer ($\alpha_i$ and $\beta_i$) via $f_i(x) = \text{BaseAct}(x)(\alpha_i - \beta_i) + \beta_i x$, allowing adaptive nonlinearity allocation across depth. Results show strong vision benchmarks (+14% on CIFAR-10), robustness to extreme depth scaling (95%+ accuracy on 75-layer MNIST), and faster convergence (24% AUC reduction), though LLM perplexity gains vanish in long-run training.

Autonomous vehicles struggle with adverse weather perception. This paper proposes LRC-WeatherNet, a lightweight fusion network combining LiDAR, RADAR, and camera via early BEV fusion and mid-level gating to classify weather conditions in real-time. The approach achieves $86.66\%$ accuracy on the MSU-4S dataset with $7.13\,\mathrm{ms}$ inference, demonstrating that adaptive multi-modal fusion outperforms unimodal baselines, though dataset limitations restrict generalization to rare weather events.