Vision-language models, speech and cross-modal systems that read, look and listen in the same forward pass.
arXiv:2608.22857v1 Announce Type: new Abstract: Vision-language models (VLMs) frequently fail at visual change reasoning, even when their vision encoders contain sufficient information. We observe th...
arXiv:2608.22872v1 Announce Type: new Abstract: Speech-based applications pass spoken queries through automatic speech recognition (ASR) before any retrieval module, so ASR errors enter the pipeline...
arXiv:2608.22316v1 Announce Type: cross Abstract: Whether an intermediate stage of modern Arabic handwriting helps or hurts historical Arabic HTR is usually decided from one implementation and one co...
The paper presents a scalable approach to harmful content moderation on social media by leveraging large language models (LLMs) for few-shot, in-context learning. Experiments across multiple LLMs show that this method outperforms proprietary baselines such as Perspective and OpenAI Moderation, as well as prior few-shot learning techniques, in detecting harmful content. The study also explores the addition of visual cues like video thumbnails to assess multimodal improvements, highlighting the advantages of LLM-based moderation for dynamic and large-scale content filtering.
FlowExtract is a pipeline designed to convert ISO 5807-standardized maintenance flowcharts into directed graphs. It separates node detection—using YOLOv8 and EasyOCR—from connectivity reconstruction, employing a novel edge detection method that traces arrowheads back to source nodes. Evaluations on industrial troubleshooting guides show high node detection accuracy and significant improvement over vision‑language model baselines for edge extraction.
The paper introduces UR$^{2}$-MLLM, an uncertainty‑aware multimodal large language model that dynamically revisits uncertain image regions during radiology report generation. It incorporates an uncertainty perception module trained on a specialized dataset, builds a multimodal reasoning trajectory with a detect‑and‑copy mechanism to guide revisits, and refines this behavior using a visual grounding reward via reinforcement learning. Experiments on MIMIC‑CXR and IU‑Xray demonstrate state‑of‑the‑art performance, underscoring the importance of visual revisit reasoning for reliable, clinically aligned reports.
Mol-JEPA is a scalable multimodal framework that learns molecular world models by using modality masking instead of suboptimal perturbations. It incorporates diverse data such as molecular structures, cellular phenotypes, binding affinities, ADMET profiles, quantum chemistry simulations, and other drug‑discovery information. Benchmarks show that the representations it learns perform strongly, highlighting the benefit of embedding biochemical context via latent‑space prediction.
LanGuSTE is a patch‑selection framework for whole slide image analysis that uses vision‑language models and large language model knowledge. It introduces Cross‑Scale Visual Prompt Tuning to align low‑resolution and high‑resolution patches, and a coarse‑to‑fine selection module that encodes only informative high‑resolution patches. Experiments show LanGuSTE cuts overall processing time to about one‑third of the baseline while matching or surpassing diagnostic performance of exhaustive and state‑of‑the‑art methods.
arXiv:2608.21864v1 Announce Type: cross Abstract: The current progress of Clinical Vision Large Language Models (C-VLLMs) has substantially improved digital diagnostics, still these frameworks often...
arXiv:2608.21500v1 Announce Type: cross Abstract: Prompt injection is listed as the \#1 threat to AI agents. When an agent accesses external data from websites, files, or emails, an attacker may inje...
arXiv:2608.22214v1 Announce Type: new Abstract: In domain-specific multimodal long documents, images and text jointly convey complex knowledge that cannot be fully captured by plain text alone. Howev...
arXiv:2608.22143v1 Announce Type: new Abstract: Qualitative mechanical problem-solving (QMPS) refers to solving qualitative problems from the mechanical domain. Qualitative problems can be solved wit...
arXiv:2608.21868v1 Announce Type: new Abstract: Depression assessment from multimodal clinical interviews requires integrating dispersed evidence from multiple symptoms into a coherent PHQ-8 profile....
arXiv:2608.22847v1 Announce Type: new Abstract: Vision-Language Models (VLMs) based GUI agents stand to benefit significantly from online reinforcement learning (RL). However, their training is bottl...
arXiv:2608.21487v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) exhibit strong zero-shot capabilities, making them an attractive solution for continual learning across diverse tasks....
arXiv:2606.01063v3 Announce Type: replace Abstract: Theory-of-Mind (ToM) reasoning enables embodied agents to understand human beliefs, goals, and intentions, but existing benchmarks mainly evaluate...
arXiv:2608.21913v1 Announce Type: new Abstract: Completing dental records from cone-beam computed tomography (CBCT) is difficult when annotation is scarce and individual clinical fields are supported...
arXiv:2608.22465v1 Announce Type: new Abstract: High-fidelity free-viewpoint video (FVV) and interactive rendering increasingly rely on explicit Gaussian representations, yet practical deployment rem...
arXiv:2608.22637v1 Announce Type: new Abstract: Recent vision-language models (VLMs) show strong capabilities in robotic perception and spatial reasoning, yet their ability to reason about complex me...
arXiv:2608.22419v1 Announce Type: cross Abstract: Query-based Vision-Language-Action (VLA) models offer low-latency inference that is attractive for bimanual robotic manipulation, but we observe that...