DIQ-H: Evaluating Hallucination Persistence in VLMs Under Temporal Visual Degradation

A novel self-supervised symbolic auto-encoder has been introduced, enabling symbolic computer vision to interpret diagrams through structured representations and logical rules. This approach contrasts with traditional pixel-based visual models by parsing diagrams into geometric primitives, enhancing machine vision's interpretability.

A comparative study has been conducted on the object counting capabilities of two multi-modal large language models, GPT-4o and GPT-5, focusing on their performance in zero-shot scenarios using only textual prompts. The evaluation was carried out on the FSC-147 and CARPK datasets, revealing that both models achieved results comparable to state-of-the-art methods, with some instances exceeding them.

A new framework called 'Look, Recite, Then Answer' has been proposed to enhance the performance of Vision-Language Models (VLMs) by generating self-generated knowledge hints. This approach aims to address the limitations of VLMs in specialized fields like precision agriculture, where reasoning-driven hallucination can hinder accurate visual perception.

A recent study has introduced FragFake, a large-scale benchmark aimed at improving the detection and localization of fine-grained AI-edited images. This initiative addresses significant challenges in current AI-generated content (AIGC) detection methods, which often fail to pinpoint where edits occur and rely on expensive pixel-level annotations. The research explores the capabilities of vision language models (VLMs) in classifying edited images and identifying specific edited regions.

A new method for Scene Graph Anticipation (SGA) has been introduced, termed Linguistic Scene Graph Anticipation (LSGA), which utilizes a language-driven framework to enhance the prediction of future scene graphs from video clips. This approach aims to improve the understanding of dynamic scenes by integrating semantic dynamics and commonsense temporal regularities, which are often difficult to extract from visual features alone.

A recent study highlights the challenges faced by vision-language models (VLMs) in factual recall, identifying a two-hop problem that involves forming entity representations from visual inputs and recalling associated knowledge. The research benchmarks 14 VLMs, revealing that 11 of them show a decline in factual recall performance compared to their large language model (LLM) counterparts.

The introduction of SpatialReasoner marks a significant advancement in spatial reasoning for large-scale 3D environments, addressing challenges faced by existing vision-language models that are limited to smaller, room-scale scenarios. This framework utilizes the H$^2$U3D dataset, which encompasses multi-floor environments and generates diverse question-answer pairs to enhance 3D scene understanding.

The introduction of the EEA framework marks a significant advancement in long video understanding, addressing challenges related to the efficient navigation of extensive visual data. EEA balances exploration and exploitation through a hierarchical tree search process, enabling the autonomous discovery of task-relevant semantic queries and the collection of closely matched video frames as semantic anchors.