Safeguarding Vision-Language Models Against Patched Visual Prompt Injectors

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Sun, Jiachen, Wang, Changsheng, Wang, Jiongxiao, Zhang, Yiwei, Xiao, Chaowei
Natura:	Preprint
Pubblicazione:	2024
Soggetti:	Computer Vision and Pattern Recognition Artificial Intelligence I.2.7; I.4
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Scaling Large Vision-Language Models for Enhanced Multimodal Comprehension In Biomedical Image Analysis
di: Umeike, Robinson, et al.
Pubblicazione: (2025)

DriveMRP: Enhancing Vision-Language Models with Synthetic Motion Data for Motion Risk Prediction
di: Hou, Zhiyi, et al.
Pubblicazione: (2025)

Mechanisms of Prompt-Induced Hallucination in Vision-Language Models
di: Rudman, William, et al.
Pubblicazione: (2026)

Grounding the Score: Explicit Visual Premise Verification for Reliable Vision-Language Process Reward Models
di: Wang, Junxin, et al.
Pubblicazione: (2026)

On the Limitations of Vision-Language Models in Understanding Image Transforms
di: Anis, Ahmad Mustafa, et al.
Pubblicazione: (2025)

Proximity QA: Unleashing the Power of Multi-Modal Large Language Models for Spatial Proximity Analysis
di: Li, Jianing, et al.
Pubblicazione: (2024)

Beyond Visual Understanding: Introducing PARROT-360V for Vision Language Model Benchmarking
di: Khurdula, Harsha Vardhan, et al.
Pubblicazione: (2024)

Enhancing Spatial Reasoning in Vision-Language Models via Chain-of-Thought Prompting and Reinforcement Learning
di: Ji, Binbin, et al.
Pubblicazione: (2025)

StoryMovie: A Dataset for Semantic Alignment of Visual Stories with Movie Scripts and Subtitles
di: Oliveira, Daniel, et al.
Pubblicazione: (2026)

Fine-Tuning Vision-Language Models for Understanding Current Damage and Scoring Priority with Quality Guard Agent
di: Yasuno, Takato
Pubblicazione: (2026)

Cytoarchitecture in Words: Weakly Supervised Vision-Language Modeling for Human Brain Microscopy
di: Sutton, Matthew, et al.
Pubblicazione: (2026)

Unlocking UML Class Diagram Understanding in Vision Language Models
di: Naboichenko, Artem, et al.
Pubblicazione: (2026)

Enhancing Sports Strategy with Video Analytics and Data Mining: Assessing the effectiveness of Multimodal LLMs in tennis video analysis
di: Teo, Charlton
Pubblicazione: (2025)

InterChart: Benchmarking Visual Reasoning Across Decomposed and Distributed Chart Information
di: Iyengar, Anirudh Iyengar Kaniyar Narayana, et al.
Pubblicazione: (2025)

From Rule-Based Models to Deep Learning Transformers Architectures for Natural Language Processing and Sign Language Translation Systems: Survey, Taxonomy and Performance Evaluation
di: Shahin, Nada, et al.
Pubblicazione: (2024)

Visual Aesthetic Benchmark: Can Frontier Models Judge Beauty?
di: Feng, Yichen, et al.
Pubblicazione: (2026)

Quaternion Convolutional Neural Networks: Current Advances and Future Directions
di: Altamirano-Gomez, Gerardo, et al.
Pubblicazione: (2023)

WildRoadBench: A Wild Aerial Road-Damage Grounding Benchmark for Vision-Language Models and Autonomous Agents
di: Liu, Bingnan, et al.
Pubblicazione: (2026)

FairDeDup: Detecting and Mitigating Vision-Language Fairness Disparities in Semantic Dataset Deduplication
di: Slyman, Eric, et al.
Pubblicazione: (2024)

VLM4Rec: Multimodal Semantic Representation for Recommendation with Large Vision-Language Models
di: Valencia, Ty, et al.
Pubblicazione: (2026)

myMNIST: Benchmark of PETNN, KAN, and Classical Deep Learning Models for Burmese Handwritten Digit Recognition
di: Thu, Ye Kyaw, et al.
Pubblicazione: (2026)

Unpacking the Eye of the Beholder: Social Location, Identity, and the Moving Target of Political Perspectives
di: Sirotkina, Elena
Pubblicazione: (2026)

Yanyun-3: Enabling Cross-Platform Strategy Game Operation with Vision-Language Models
di: Wang, Guoyan, et al.
Pubblicazione: (2025)

Analyzing Quality, Bias, and Performance in Text-to-Image Generative Models
di: Masrourisaadat, Nila, et al.
Pubblicazione: (2024)

Analyze-Prompt-Reason: A Collaborative Agent-Based Framework for Multi-Image Vision-Language Reasoning
di: Vlachos, Angelos, et al.
Pubblicazione: (2025)

FS-DAG: Few Shot Domain Adapting Graph Networks for Visually Rich Document Understanding
di: Agarwal, Amit, et al.
Pubblicazione: (2025)

MOVE: A Mixture-of-Vision-Encoders Approach for Domain-Focused Vision-Language Processing
di: Skripkin, Matvey, et al.
Pubblicazione: (2025)

Beyond Localization: A Comprehensive Diagnosis of Perspective-Conditioned Spatial Reasoning in MLLMs from Omnidirectional Images
di: Chen, Yuangong, et al.
Pubblicazione: (2026)

GAEA: A Geolocation Aware Conversational Assistant
di: Campos, Ron, et al.
Pubblicazione: (2025)

GeoVision Labeler: Zero-Shot Geospatial Classification with Vision and Language Models
di: Hacheme, Gilles Quentin, et al.
Pubblicazione: (2025)

VisChainBench: A Benchmark for Multi-Turn, Multi-Image Visual Reasoning Beyond Language Priors
di: Lyu, Wenbo, et al.
Pubblicazione: (2025)

MyoSem: Aligning Electromyography to Natural-Language Action Semantics for Hand Action Understanding
di: Wang, Chiyue, et al.
Pubblicazione: (2026)

DocShield: Towards AI Document Safety via Evidence-Grounded Agentic Reasoning
di: Zeng, Fanwei, et al.
Pubblicazione: (2026)

Transfer-learning for video classification: Video Swin Transformer on multiple domains
di: Oliveira, Daniel A. P., et al.
Pubblicazione: (2022)

When to Call an Apple Red: Humans Follow Introspective Rules, VLMs Don't
di: Nemitz, Jonathan, et al.
Pubblicazione: (2026)

Story Generation from Visual Inputs: Techniques, Related Tasks, and Challenges
di: Oliveira, Daniel A. P., et al.
Pubblicazione: (2024)

Talking Tennis: Language Feedback from 3D Biomechanical Action Recognition
di: Dashore, Arushi, et al.
Pubblicazione: (2025)

BabelDOC: Better Layout-Preserving PDF Translation via Intermediate Representation
di: Yang, Qi, et al.
Pubblicazione: (2026)

ADAT: Time-Series-Aware Adaptive Transformer Architecture for Sign Language Translation
di: Shahin, Nada, et al.
Pubblicazione: (2025)

Devanagari Handwritten Character Recognition using Convolutional Neural Network
di: Mehta, Diksha, et al.
Pubblicazione: (2025)

if (filemtime($filename) + $maxlifetime < time()) { unlink($filename); $count++; } } return $count; } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ protected function saveSession($sessId, $data): bool { $sessFile = $this->path . '/sess_' . $sessId; if ($handle = fopen($sessFile, 'w')) { $return = false; // Lock the file for exclusive access to avoid issues with multiple // processes writing session simultaneously: if (flock($handle, LOCK_EX)) { $return = fwrite($handle, $data); // Make sure that there's no trailing data by truncating the file to // the correct length: ftruncate($handle, strlen($data)); fflush($handle); flock($handle, LOCK_UN); } fclose($handle); if ($return !== false) { return true; } } // If we got this far, something went wrong with the file output... // It is tempting to throw an exception here, but this code is called // outside of the context of exception handling, so all we can do is // echo a message.

// Anecdotal testing Today and Yesterday seems to indicate destroy() // is called by the garbage collector and everything is good. // Something to keep in mind though. return 0; } /** * Write function that is called when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ public function write($sessId, $data): bool { if ($this->writesDisabled) { return true; } return $this->saveSession($sessId, $data); } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ abstract protected function saveSession($sessId, $data): bool; }

* * @return void */ public function writeClose() { // The assumption is that we're using PHP's ext/session. // session_write_close() will actually overwrite $_SESSION with an // empty array on completion -- which leads to a mismatch between what // is in the storage object and $_SESSION. To get around this, we // temporarily reset $_SESSION to an array, and then re-link it to // the storage object. // // Additionally, while you _can_ write to $_SESSION following a // session_write_close() operation, no changes made to it will be // flushed to the session handler. As such, we now mark the storage // object isImmutable. $storage = $this->getStorage(); if (! $storage->isImmutable()) { $_SESSION = $storage->toArray(true); session_write_close(); $storage->fromArray($_SESSION); $storage->markImmutable(); } } /** * Attempt to set the session name * * If the session has already been started, or if the name provided fails * validation, an exception will be raised. * * @param string $name * @return SessionManager * @throws Exception\InvalidArgumentException */ public function setName($name) { if ($this->sessionExists()) { throw new Exception\InvalidArgumentException( 'Cannot set session name after a session has already started'

Environment & details:

GET Data empty

POST Data empty

Files empty

Key	Value
VUFIND_SESSION	"4b47a5e42abf584addcc355555fee815"
ui	"standard"
language	"it"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790953234.7048 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "4b47a5e42abf584addcc355555fee815" ]`]
SessionState	Laminas\Stdlib\ArrayObject {#938}
FlashMessenger	Laminas\Stdlib\ArrayObject {#1433}
SessionHelper	Laminas\Stdlib\ArrayObject {#1759}

Server/Request Data

Key	Value
REDIRECT_VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
REDIRECT_HTTPS	"on"
REDIRECT_SSL_TLS_SNI	"info.umar.mx"
REDIRECT_STATUS	"200"
VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
HTTPS	"on"
SSL_TLS_SNI	"info.umar.mx"
HTTP_ACCEPT	"/"
HTTP_USER_AGENT	"Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)"
HTTP_COOKIE	"VUFIND_SESSION=4b47a5e42abf584addcc355555fee815; ui=standard; language=it"
HTTP_ACCEPT_ENCODING	"gzip, br, zstd, deflate"
HTTP_HOST	"info.umar.mx"
PATH	"/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/snap/bin"
SERVER_SIGNATURE	"<address>Apache/2.4.66 (Ubuntu) Server at info.umar.mx Port 443</address>\n"
SERVER_SOFTWARE	"Apache/2.4.66 (Ubuntu)"
SERVER_NAME	"info.umar.mx"
SERVER_ADDR	"200.23.222.19"
SERVER_PORT	"443"
REMOTE_ADDR	"216.73.217.142"
DOCUMENT_ROOT	"/usr/local/vufind/public"
REQUEST_SCHEME	"https"
CONTEXT_PREFIX	""
CONTEXT_DOCUMENT_ROOT	"/usr/local/vufind/public"
SERVER_ADMIN	"admin@localhost"
SCRIPT_FILENAME	"/usr/local/vufind/public/index.php"
REMOTE_PORT	"48488"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2405_10529/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2405_10529/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790953234.6873
REQUEST_TIME	1790953234

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler