DeepImageSearch: Benchmarking Multimodal Agents for Context-Aware Image Retrieval in Visual Histories

Existing multimodal retrieval systems excel at semantic matching but implicitly assume that query-image relevance can be measured in isolation. This paradigm overlooks the rich dependencies inherent in realistic visual streams, where information is distributed across temporal sequences rather than confined to single snapshots. To bridge this gap, we introduce DeepImageSearch, a novel agentic paradigm that reformulates image retrieval as an autonomous exploration task. Models must plan and perform multi-step reasoning over raw visual histories to locate targets based on implicit contextual cues. We construct DISBench, a challenging benchmark built on interconnected visual data. To address the scalability challenge of creating context-dependent queries, we propose a human-model collaborative pipeline that employs vision-language models to mine latent spatiotemporal associations, effectively offloading intensive context discovery before human verification. Furthermore, we build a robust baseline using a modular agent framework equipped with fine-grained tools and a dual-memory system for long-horizon navigation. Extensive experiments demonstrate that DISBench poses significant challenges to state-of-the-art models, highlighting the necessity of incorporating agentic reasoning into next-generation retrieval systems.

OSCAR: Optimization-Steered Agentic Planning for Composed Image Retrieval

Composed image retrieval (CIR) requires complex reasoning over heterogeneous visual and textual constraints. Existing approaches largely fall into two paradigms: unified embedding retrieval, which suffers from single-model myopia, and heuristic agentic retrieval, which is limited by suboptimal, trial-and-error orchestration. To this end, we propose OSCAR, an optimization-steered agentic planning framework for composed image retrieval. We are the first to reformulate agentic CIR from a heuristic search process into a principled trajectory optimization problem. Instead of relying on heuristic trial-and-error exploration, OSCAR employs a novel offline-online paradigm. In the offline phase, we model CIR via atomic retrieval selection and composition as a two-stage mixed-integer programming problem, mathematically deriving optimal trajectories that maximize ground-truth coverage for training samples via rigorous boolean set operations. These trajectories are then stored in a golden library to serve as in-context demonstrations for online steering of VLM planner at online inference time. Extensive experiments on three public benchmarks and a private industrial benchmark show that OSCAR consistently outperforms SOTA baselines. Notably, it achieves superior performance using only 10% of training data, demonstrating strong generalization of planning logic rather than dataset-specific memorization.

\textsc{NaVIDA}: Vision-Language Navigation with Inverse Dynamics Augmentation

Vision-and-Language Navigation (VLN) requires agents to interpret natural language instructions and act coherently in visually rich environments. However, most existing methods rely on reactive state-action mappings without explicitly modeling how actions causally transform subsequent visual observations. Lacking such vision-action causality, agents cannot anticipate the visual changes induced by its own actions, leading to unstable behaviors, weak generalization, and cumulative error along trajectory. To address these issues, we introduce \textsc{NaVIDA} (\textbf{Nav}igation with \textbf{I}nverse \textbf{D}ynamics \textbf{A}ugmentation), a unified VLN framework that couples policy learning with action-grounded visual dynamics and adaptive execution. \textsc{NaVIDA} augments training with chunk-based inverse-dynamics supervision to learn causal relationship between visual changes and corresponding actions. To structure this supervision and extend the effective planning range, \textsc{NaVIDA} employs hierarchical probabilistic action chunking (HPAC), which organizes trajectories into multi-step chunks and provides discriminative, longer-range visual-change cues. To further curb error accumulation and stabilize behavior at inference, an entropy-guided mechanism adaptively sets the execution horizon of action chunks. Extensive experiments show that \textsc{NaVIDA} achieves superior navigation performance compared to state-of-the-art methods with fewer parameters (3B vs. 8B). Real-world robot evaluations further validate the practical feasibility and effectiveness of our approach. Code and data will be available upon acceptance.

ActionFlow: A Pipelined Action Acceleration for Vision Language Models on Edge

Vision-Language-Action (VLA) models have emerged as a unified paradigm for robotic perception and control, enabling emergent generalization and long-horizon task execution. However, their deployment in dynamic, real-world environments is severely hin dered by high inference latency. While smooth robotic interaction requires control frequencies of 20 to 30 Hz, current VLA models typi cally operate at only 3-5 Hz on edge devices due to the memory bound nature of autoregressive decoding. Existing optimizations often require extensive retraining or compromise model accuracy. To bridge this gap, we introduce ActionFlow, a system-level inference framework tailored for resource-constrained edge plat forms. At the core of ActionFlow is a Cross-Request Pipelin ing strategy, a novel scheduler that redefines VLA inference as a macro-pipeline of micro-requests. The strategy intelligently batches memory-bound Decode phases with compute-bound Prefill phases across continuous time steps to maximize hardware utilization. Furthermore, to support this scheduling, we propose a Cross Request State Packed Forward operator and a Unified KV Ring Buffer, which fuse fragmented memory operations into efficient dense computations. Experimental results demonstrate that ActionFlow achieves a 2.55x improvement in FPS on the OpenVLA-7B model without retraining, enabling real-time dy namic manipulation on edge hardware. Our work is available at https://anonymous.4open.science/r/ActionFlow-1D47.

TimeLens: Rethinking Video Temporal Grounding with Multimodal LLMs

This paper does not introduce a novel method but instead establishes a straightforward, incremental, yet essential baseline for video temporal grounding (VTG), a core capability in video understanding. While multimodal large language models (MLLMs) excel at various video understanding tasks, the recipes for optimizing them for VTG remain under-explored. In this paper, we present TimeLens, a systematic investigation into building MLLMs with strong VTG ability, along two primary dimensions: data quality and algorithmic design. We first expose critical quality issues in existing VTG benchmarks and introduce TimeLens-Bench, comprising meticulously re-annotated versions of three popular benchmarks with strict quality criteria. Our analysis reveals dramatic model re-rankings compared to legacy benchmarks, confirming the unreliability of prior evaluation standards. We also address noisy training data through an automated re-annotation pipeline, yielding TimeLens-100K, a large-scale, high-quality training dataset. Building on our data foundation, we conduct in-depth explorations of algorithmic design principles, yielding a series of meaningful insights and effective yet efficient practices. These include interleaved textual encoding for time representation, a thinking-free reinforcement learning with verifiable rewards (RLVR) approach as the training paradigm, and carefully designed recipes for RLVR training. These efforts culminate in TimeLens models, a family of MLLMs with state-of-the-art VTG performance among open-source models and even surpass proprietary models such as GPT-5 and Gemini-2.5-Flash. All codes, data, and models will be released to facilitate future research.

ARC-Chapter: Structuring Hour-Long Videos into Navigable Chapters and Hierarchical Summaries

The proliferation of hour-long videos (e.g., lectures, podcasts, documentaries) has intensified demand for efficient content structuring. However, existing approaches are constrained by small-scale training with annotations that are typical short and coarse, restricting generalization to nuanced transitions in long videos. We introduce ARC-Chapter, the first large-scale video chaptering model trained on over million-level long video chapters, featuring bilingual, temporally grounded, and hierarchical chapter annotations. To achieve this goal, we curated a bilingual English-Chinese chapter dataset via a structured pipeline that unifies ASR transcripts, scene texts, visual captions into multi-level annotations, from short title to long summaries. We demonstrate clear performance improvements with data scaling, both in data volume and label intensity. Moreover, we design a new evaluation metric termed GRACE, which incorporates many-to-one segment overlaps and semantic similarity, better reflecting real-world chaptering flexibility. Extensive experiments demonstrate that ARC-Chapter establishes a new state-of-the-art by a significant margin, outperforming the previous best by 14.0% in F1 score and 11.3% in SODA score. Moreover, ARC-Chapter shows excellent transferability, improving the state-of-the-art on downstream tasks like dense video captioning on YouCook2.

FLYINGTRUST: A Benchmark for Quadrotor Navigation Across Scenarios and Vehicles

Visual navigation algorithms for quadrotors often exhibit a large variation in performance when transferred across different vehicle platforms and scene geometries, which increases the cost and risk of field deployment. To support systematic early-stage evaluation, we introduce FLYINGTRUST, a high-fidelity, configurable benchmarking framework that measures how platform kinodynamics and scenario structure jointly affect navigation robustness. FLYINGTRUST models vehicle capability with two compact, physically interpretable indicators: maximum thrust-to-weight ratio and axis-wise maximum angular acceleration. The benchmark pairs a diverse scenario library with a heterogeneous set of real and virtual platforms and prescribes a standardized evaluation protocol together with a composite scoring method that balances scenario importance, platform importance and performance stability. We use FLYINGTRUST to compare representative optimization-based and learning-based navigation approaches under identical conditions, performing repeated trials per platform-scenario combination and reporting uncertainty-aware metrics. The results reveal systematic patterns: navigation success depends predictably on platform capability and scene geometry, and different algorithms exhibit distinct preferences and failure modes across the evaluated conditions. These observations highlight the practical necessity of incorporating both platform capability and scenario structure into algorithm design, evaluation, and selection, and they motivate future work on methods that remain robust across diverse platforms and scenarios.

UltraHiT: A Hierarchical Transformer Architecture for Generalizable Internal Carotid Artery Robotic Ultrasonography

Carotid ultrasound is crucial for the assessment of cerebrovascular health, particularly the internal carotid artery (ICA). While previous research has explored automating carotid ultrasound, none has tackled the challenging ICA. This is primarily due to its deep location, tortuous course, and significant individual variations, which greatly increase scanning complexity. To address this, we propose a Hierarchical Transformer-based decision architecture, namely UltraHiT, that integrates high-level variation assessment with low-level action decision. Our motivation stems from conceptualizing individual vascular structures as morphological variations derived from a standard vascular model. The high-level module identifies variation and switches between two low-level modules: an adaptive corrector for variations, or a standard executor for normal cases. Specifically, both the high-level module and the adaptive corrector are implemented as causal transformers that generate predictions based on the historical scanning sequence. To ensure generalizability, we collected the first large-scale ICA scanning dataset comprising 164 trajectories and 72K samples from 28 subjects of both genders. Based on the above innovations, our approach achieves a 95% success rate in locating the ICA on unseen individuals, outperforming baselines and demonstrating its effectiveness. Our code will be released after acceptance.

Predicting person-level injury severity using crash narratives: A balanced approach with roadway classification and natural language process techniques

Predicting injuries and fatalities in traffic crashes plays a critical role in enhancing road safety, improving emergency response, and guiding public health interventions. This study investigates the added value of unstructured crash narratives (written by police officers at the scene) when combined with structured crash data to predict injury severity. Two widely used Natural Language Processing (NLP) techniques, Term Frequency-Inverse Document Frequency (TF-IDF) and Word2Vec, were employed to extract semantic meaning from the narratives, and their effectiveness was compared. To address the challenge of class imbalance, a K-Nearest Neighbors-based oversampling method was applied to the training data prior to modeling. The dataset consists of crash records from Kentucky spanning 2019 to 2023. To account for roadway heterogeneity, three road classification schemes were used: (1) eight detailed functional classes (e.g., Urban Two-Lane, Rural Interstate, Urban Multilane Divided), (2) four broader paired categories (e.g., Urban vs. Rural, Freeway vs. Non-Freeway), and (3) a unified dataset without classification. A total of 102 machine learning models were developed by combining structured features and narrative-based features using the two NLP techniques alongside three ensemble algorithms: XGBoost, Random Forest, and AdaBoost. Results demonstrate that models incorporating narrative data consistently outperform those relying solely on structured data. Among all combinations, TF-IDF coupled with XGBoost yielded the most accurate predictions in most subgroups. The findings highlight the power of integrating textual and structured crash information to enhance person-level injury prediction. This work offers a practical and adaptable framework for transportation safety professionals to improve crash severity modeling, guide policy decisions, and design more effective countermeasures.

CVBench: Evaluating Cross-Video Synergies for Complex Multimodal Understanding and Reasoning

While multimodal large language models (MLLMs) exhibit strong performance on single-video tasks (e.g., video question answering), their ability across multiple videos remains critically underexplored. However, this capability is essential for real-world applications, including multi-camera surveillance and cross-video procedural learning. To bridge this gap, we present CVBench, the first comprehensive benchmark designed to assess cross-video relational reasoning rigorously. CVBench comprises 1,000 question-answer pairs spanning three hierarchical tiers: cross-video object association (identifying shared entities), cross-video event association (linking temporal or causal event chains), and cross-video complex reasoning (integrating commonsense and domain knowledge). Built from five domain-diverse video clusters (e.g., sports, life records), the benchmark challenges models to synthesise information across dynamic visual contexts. Extensive evaluation of 10+ leading MLLMs (including GPT-4o, Gemini-2.0-flash, Qwen2.5-VL) under zero-shot or chain-of-thought prompting paradigms. Key findings reveal stark performance gaps: even top models, such as GPT-4o, achieve only 60% accuracy on causal reasoning tasks, compared to the 91% accuracy of human performance. Crucially, our analysis reveals fundamental bottlenecks inherent in current MLLM architectures, notably deficient inter-video context retention and poor disambiguation of overlapping entities. CVBench establishes a rigorous framework for diagnosing and advancing multi-video reasoning, offering architectural insights for next-generation MLLMs. The data and evaluation code are available at https://github.com/Hokhim2/CVBench.