FineBench: Benchmarking and Enhancing Vision-Language Models for Fine-grained Human Activity Understanding

Vision-Language Models (VLMs) have demonstrated remarkable capabilities in general video understanding, yet they often struggle with the fine-grained comprehension crucial for real-world applications requiring nuanced interpretation of human actions and interactions. While some recent human-centric benchmarks evaluate aspects of model behaviour such as fairness/ethics, emotion perception, and broader human-centric metrics, they do not combine long-form videos, very dense QA coverage, and frame-level spatial/temporal grounding at scale. To bridge this gap, we introduce FineBench, a human-centric video question answering (VQA) benchmark specifically designed to assess fine-grained understanding. FineBench comprises 199,420 multiple-choice QA pairs densely annotated across 64 long-form videos (15 minutes each), focusing on detailed person movement, person interaction, and object manipulation, including compositional actions. Our extensive evaluation reveals that while proprietary models like GPT-5 achieve respectable performance, current open-source VLMs significantly underperform, struggling particularly with spatial reasoning in multi-person scenes and distinguishing subtle differences in human movements and interactions. To address these identified weaknesses, we propose FineAgent, a modular framework that enhances VLMs by leveraging a Localizer and a Descriptor. Experiments show that FineAgent consistently improves the performance of various open VLMs on FineBench. FineBench provides a rigorous testbed for future research into fine-grained human-centric video understanding, while FineAgent offers a practical approach to enhance such reasoning in current VLMs. Project page and code at https://joslefaure.github.io/assets/html/finebench.html.

SPATIOROUTE: Dynamic Prompt Routing for Zero-Shot Spatial Reasoning

Spatial question answering over egocentric video is a challenging task that requires Vision-Language Models (VLMs) to reason about 3D object positions, scene affordances, and directional relationships, particularly in the zero-shot setting where no task-specific fine-tuning is available. We introduce SpatioRoute, a dynamic prompt generation approach that routes each incoming question to a semantically tailored prompt template -- without any additional training, fine-tuning, or 3D sensor input. SpatioRoute operates in two complementary modes: SpatioRoute-R, a rule-based router that deterministically maps question typologies (e.g., What, Is, How, Can, Which) to specialized prompt templates; and SpatioRoute-L, an LLM-driven approach that generates task-specific prompts from the question and situational context alone, with no video input at routing time. We evaluate SpatioRoute on the SQA3D benchmark across VLMs spanning model families. SpatioRoute achieves consistent overall accuracy gains up to 5% over fixed prompt baselines, establishing a new state-of-the-art for zero-shot video-only spatial VQA without requiring 3D point-cloud inputs. As an additional finding, we observe that Chain-of-Thought (CoT) prompting, implemented via the Think it Twice architecture, consistently degrades performance in this setting on Qwen series models, confirming that question-aware routing is more effective than uniform reasoning instructions for spatial video understanding.

SceneFunRI: Reasoning the Invisible for Task-Driven Functional Object Localization

In real-world scenes, target objects may reside in regions that are not visible. While humans can often infer the locations of occluded objects from context and commonsense knowledge, this capability remains a major challenge for vision-language models (VLMs). To address this gap, we introduce SceneFunRI, a benchmark for Reasoning the Invisible. Based on the SceneFun3D dataset, SceneFunRI formulates the task as a 2D spatial reasoning problem via a semi-automatic pipeline and comprises 855 instances. It requires models to infer the locations of invisible functional objects from task instructions and commonsense reasoning. The strongest baseline model (Gemini 3 Flash) only achieves an CAcc@75 of 15.20, an mIoU of 0.74, and a Dist of 28.65. We group our prompting analysis into three categories: Strong Instruction Prompting, Reasoning-based Prompting, and Spatial Process of Elimination (SPoE). These findings indicate that invisible-region reasoning remains an unstable capability in current VLMs, motivating future work on models that more tightly integrate task intent, commonsense priors, spatial grounding, and uncertainty-aware search.

VLN-NF: Feasibility-Aware Vision-and-Language Navigation with False-Premise Instructions

Conventional Vision-and-Language Navigation (VLN) benchmarks assume instructions are feasible and the referenced target exists, leaving agents ill-equipped to handle false-premise goals. We introduce VLN-NF, a benchmark with false-premise instructions where the target is absent from the specified room and agents must navigate, gather evidence through in-room exploration, and explicitly output NOT-FOUND. VLN-NF is constructed via a scalable pipeline that rewrites VLN instructions using an LLM and verifies target absence with a VLM, producing plausible yet factually incorrect goals. We further propose REV-SPL to jointly evaluate room reaching, exploration coverage, and decision correctness. To address this challenge, we present ROAM, a two-stage hybrid that combines supervised room-level navigation with LLM/VLM-driven in-room exploration guided by a free-space clearance prior. ROAM achieves the best REV-SPL among compared methods, while baselines often under-explore and terminate prematurely under unreliable instructions. VLN-NF project page can be found at https://vln-nf.github.io/.

Affordance-Guided Coarse-to-Fine Exploration for Base Placement in Open-Vocabulary Mobile Manipulation

In open-vocabulary mobile manipulation (OVMM), task success often hinges on the selection of an appropriate base placement for the robot. Existing approaches typically navigate to proximity-based regions without considering affordances, resulting in frequent manipulation failures. We propose Affordance-Guided Coarse-to-Fine Exploration, a zero-shot framework for base placement that integrates semantic understanding from vision-language models (VLMs) with geometric feasibility through an iterative optimization process. Our method constructs cross-modal representations, namely Affordance RGB and Obstacle Map+, to align semantics with spatial context. This enables reasoning that extends beyond the egocentric limitations of RGB perception. To ensure interaction is guided by task-relevant affordances, we leverage coarse semantic priors from VLMs to guide the search toward task-relevant regions and refine placements with geometric constraints, thereby reducing the risk of convergence to local optima. Evaluated on five diverse open-vocabulary mobile manipulation tasks, our system achieves an 85% success rate, significantly outperforming classical geometric planners and VLM-based methods. This demonstrates the promise of affordance-aware and multimodal reasoning for generalizable, instruction-conditioned planning in OVMM.

MovieCORE: COgnitive REasoning in Movies

This paper introduces MovieCORE, a novel video question answering (VQA) dataset designed to probe deeper cognitive understanding of movie content. Unlike existing datasets that focus on surface-level comprehension, MovieCORE emphasizes questions that engage System-2 thinking while remaining specific to the video material. We present an innovative agentic brainstorming approach, utilizing multiple large language models (LLMs) as thought agents to generate and refine high-quality question-answer pairs. To evaluate dataset quality, we develop a set of cognitive tests assessing depth, thought-provocation potential, and syntactic complexity. We also propose a comprehensive evaluation scheme for assessing VQA model performance on deeper cognitive tasks. To address the limitations of existing video-language models (VLMs), we introduce an agentic enhancement module, Agentic Choice Enhancement (ACE), which improves model reasoning capabilities post-training by up to 25%. Our work contributes to advancing movie understanding in AI systems and provides valuable insights into the capabilities and limitations of current VQA models when faced with more challenging, nuanced questions about cinematic content. Our project page, dataset and code can be found at https://joslefaure.github.io/assets/html/moviecore.html.

Improving Generalization Ability for 3D Object Detection by Learning Sparsity-invariant Features

In autonomous driving, 3D object detection is essential for accurately identifying and tracking objects. Despite the continuous development of various technologies for this task, a significant drawback is observed in most of them-they experience substantial performance degradation when detecting objects in unseen domains. In this paper, we propose a method to improve the generalization ability for 3D object detection on a single domain. We primarily focus on generalizing from a single source domain to target domains with distinct sensor configurations and scene distributions. To learn sparsity-invariant features from a single source domain, we selectively subsample the source data to a specific beam, using confidence scores determined by the current detector to identify the density that holds utmost importance for the detector. Subsequently, we employ the teacher-student framework to align the Bird's Eye View (BEV) features for different point clouds densities. We also utilize feature content alignment (FCA) and graph-based embedding relationship alignment (GERA) to instruct the detector to be domain-agnostic. Extensive experiments demonstrate that our method exhibits superior generalization capabilities compared to other baselines. Furthermore, our approach even outperforms certain domain adaptation methods that can access to the target domain data.

Leveraging Content and Context Cues for Low-Light Image Enhancement

Low-light conditions have an adverse impact on machine cognition, limiting the performance of computer vision systems in real life. Since low-light data is limited and difficult to annotate, we focus on image processing to enhance low-light images and improve the performance of any downstream task model, instead of fine-tuning each of the models which can be prohibitively expensive. We propose to improve the existing zero-reference low-light enhancement by leveraging the CLIP model to capture image prior and for semantic guidance. Specifically, we propose a data augmentation strategy to learn an image prior via prompt learning, based on image sampling, to learn the image prior without any need for paired or unpaired normal-light data. Next, we propose a semantic guidance strategy that maximally takes advantage of existing low-light annotation by introducing both content and context cues about the image training patches. We experimentally show, in a qualitative study, that the proposed prior and semantic guidance help to improve the overall image contrast and hue, as well as improve background-foreground discrimination, resulting in reduced over-saturation and noise over-amplification, common in related zero-reference methods. As we target machine cognition, rather than rely on assuming the correlation between human perception and downstream task performance, we conduct and present an ablation study and comparison with related zero-reference methods in terms of task-based performance across many low-light datasets, including image classification, object and face detection, showing the effectiveness of our proposed method.

Attention Tracker: Detecting Prompt Injection Attacks in LLMs

Large Language Models (LLMs) have revolutionized various domains but remain vulnerable to prompt injection attacks, where malicious inputs manipulate the model into ignoring original instructions and executing designated action. In this paper, we investigate the underlying mechanisms of these attacks by analyzing the attention patterns within LLMs. We introduce the concept of the distraction effect, where specific attention heads, termed important heads, shift focus from the original instruction to the injected instruction. Building on this discovery, we propose Attention Tracker, a training-free detection method that tracks attention patterns on instruction to detect prompt injection attacks without the need for additional LLM inference. Our method generalizes effectively across diverse models, datasets, and attack types, showing an AUROC improvement of up to 10.0% over existing methods, and performs well even on small LLMs. We demonstrate the robustness of our approach through extensive evaluations and provide insights into safeguarding LLM-integrated systems from prompt injection vulnerabilities.

Unveiling Narrative Reasoning Limits of Large Language Models with Trope in Movie Synopses

Large language models (LLMs) equipped with chain-of-thoughts (CoT) prompting have shown significant multi-step reasoning capabilities in factual content like mathematics, commonsense, and logic. However, their performance in narrative reasoning, which demands greater abstraction capabilities, remains unexplored. This study utilizes tropes in movie synopses to assess the abstract reasoning abilities of state-of-the-art LLMs and uncovers their low performance. We introduce a trope-wise querying approach to address these challenges and boost the F1 score by 11.8 points. Moreover, while prior studies suggest that CoT enhances multi-step reasoning, this study shows CoT can cause hallucinations in narrative content, reducing GPT-4's performance. We also introduce an Adversarial Injection method to embed trope-related text tokens into movie synopses without explicit tropes, revealing CoT's heightened sensitivity to such injections. Our comprehensive analysis provides insights for future research directions.