CVE-2013-0139

The Arecont Vision AV1355DN MegaDome camera allows remote attackers to cause a denial of service video-capture outage via a packet to UDP port 69...

BadVLA: Towards Backdoor Attacks on Vision-Language-Action Models Via Objective-Decoupled Optimization

Vision-Language-Action VLA models have advanced robotic control by enabling end-to-end decision-making directly from multimodal inputs. However, their tightly coupled architectures expose novel security vulnerabilities. Unlike traditional adversarial perturbations, backdoor attacks represent a...

Blind Spot Navigation: Evolutionary Discovery of Sensitive Semantic Concepts for LVLMs

Whitepaper called Blind Spot Navigation: Evolutionary Discovery Of Sensitive Semantic Concepts For LVLMs...

Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study

Rapid deployment of vision-language models VLMs magnifies safety risks, yet most evaluations rely on artificial images. This study asks: How safe are current VLMs when confronted with meme images that ordinary users share? To investigate this question, we introduce MemeSafetyBench, a...

FragFake: a Dataset for Fine-Grained Detection of Edited Images with Vision Language Models

Fine-grained edited image detection of localized edits in images is crucial for assessing content authenticity, especially given that modern diffusion models and image editing methods can produce highly realistic manipulations. However, this domain faces three challenges: 1 Binary classifiers yie...

Beyond Text: Unveiling Privacy Vulnerabilities in Multi-Modal Retrieval-Augmented Generation

Multimodal Retrieval-Augmented Generation MRAG systems enhance LMMs by integrating external multimodal databases, but introduce unexplored privacy vulnerabilities. While text-based RAG privacy risks have been studied, multimodal data presents unique challenges. We provide the first systematic...

SUSE CVE-2025-31215

The issue was addressed with improved checks. This issue is fixed in Safari 18.5, iOS 18.5 and iPadOS 18.5, iPadOS 17.7.7, macOS Sequoia 15.5, tvOS 18.5, visionOS 2.5, watchOS 11.5. Processing maliciously crafted web content may lead to an unexpected process crash...

Friday Squid Blogging: Pet Squid Simulation

From Hackaday.com, this is a neural network simulation of a pet squid. Autonomous Behavior: The squid moves autonomously, making decisions based on his current state hunger, sleepiness, etc.. Implements a vision cone for food detection, simulating realistic foraging behavior. Neural network can...

GuardReasoner-VL: Safeguarding VLMs Via Reinforced Reasoning

To enhance the safety of VLMs, this paper introduces a novel reasoning-based VLM guard model dubbed GuardReasoner-VL. The core idea is to incentivize the guard model to deliberatively reason before making moderation decisions via online RL. First, we construct GuardReasoner-VLTrain, a reasoning...

Diverging Towards Hallucination: Detection of Failures in Vision-Language Models Via Multi-Token Aggregation

Vision-language models VLMs now rival human performance on many multimodal tasks, yet they still hallucinate objects or generate unsafe text. Current hallucination detectors, e.g., single-token linear probing SLP and PTrue, typically analyze only the logit of the first generated token or just its...

SQL Injection Vulnerability in Remote Medical Comprehensive Service Platform of Beijing Divine Vision Han Technology Co.

Founded in 2010, Beijing Divine Vision Han Technology Co., Ltd. is an information technology product and solution provider focusing on the fields of intelligent hospitals, commercial intelligent screens, intelligent education and so on. A SQL injection vulnerability exists in the remote medical...

CVE-2025-47682

Improper Neutralization of Special Elements used in an SQL Command 'SQL Injection' vulnerability in Cozy Vision SMS Alert Order Notifications sms-alert allows SQL Injection.This issue affects SMS Alert Order Notifications: from n/a through = 3.8.1...

The vulnerability of the RepairKit component in iPadOS, iOS, and visionOS allows attackers to compromise the confidentiality of protected information.

The vulnerability of the RepairKit component in iPadOS, iOS, and visionOS is related to insufficient validation of input data. Exploiting this vulnerability could allow attackers to compromise the confidentiality of protected information...

The vulnerability of the Safari browser on operating systems such as visionOS, macOS, iOS, and iPadOS allows a hacker to execute arbitrary code.

The vulnerability of the Safari browser in operating systems such as visionOS, macOS, iOS, and iPadOS is related to buffer overflow in dynamic memory. Exploiting this vulnerability allows a remote attacker to execute arbitrary code...

The vulnerability of the Power Management component in operating systems such as visionOS, MacOS, tvOS, iPadOS, and iOS allows attackers to trigger service failures.

The vulnerability of the Power Management component in operating systems such as visionOS, MacOS, tvOS, iPadOS, and iOS is related to lack of access control. Exploiting this vulnerability can allow an attacker to cause service failures...

The vulnerability of the CoreMedia component in operating systems such as macOS, tvOS, iPadOS, iOS, visionOS, and watchOS allows a hacker to trigger a service failure.

The vulnerability of the CoreMedia component in operating systems such as macOS, tvOS, iPadOS, iOS, visionOS, and watchOS relates to the ability to exploit memory after it is freed. Exploiting this vulnerability could allow an attacker to cause a service failure...

Geo Vision EoL Devices Improper Neutralization of Special Elements used in an OS Command (CVE-2024-6047)

Certain EOL GeoVision devices fail to properly filter user input for the specific functionality. Unauthenticated remote attackers can exploit this vulnerability to inject and execute arbitrary system commands on the device. This plugin only works with Tenable.ot. Please visit...

Removing Watermarks with Partial Regeneration Using Semantic Information

As AI-generated imagery becomes ubiquitous, invisible watermarks have emerged as a primary line of defense for copyright and provenance. The newest watermarking schemes embed semantic signals - content-aware patterns that are designed to survive common image manipulations - yet their true...

CVE-2025-31217

The issue was addressed with improved input validation. This issue is fixed in watchOS 11.5, tvOS 18.5, iPadOS 17.7.7, iOS 18.5 and iPadOS 18.5, macOS Sequoia 15.5, visionOS 2.5, Safari 18.5. Processing maliciously crafted web content may lead to an unexpected Safari crash...

DEBIAN-CVE-2025-31215

The issue was addressed with improved checks. This issue is fixed in Safari 18.5, iOS 18.5 and iPadOS 18.5, iPadOS 17.7.7, macOS Sequoia 15.5, tvOS 18.5, visionOS 2.5, watchOS 11.5. Processing maliciously crafted web content may lead to an unexpected process crash...