REVIEW 5 cited by
COSMOS: Catching Out-of-Context Misinformation with Self-Supervised Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Despite the recent attention to DeepFakes, one of the most prevalent ways to mislead audiences on social media is the use of unaltered images in a new but false context. To address these challenges and support fact-checkers, we propose a new method that automatically detects out-of-context image and text pairs. Our key insight is to leverage the grounding of image with text to distinguish out-of-context scenarios that cannot be disambiguated with language alone. We propose a self-supervised training strategy where we only need a set of captioned images. At train time, our method learns to selectively align individual objects in an image with textual claims, without explicit supervision. At test time, we check if both captions correspond to the same object(s) in the image but are semantically different, which allows us to make fairly accurate out-of-context predictions. Our method achieves 85% out-of-context detection accuracy. To facilitate benchmarking of this task, we create a large-scale dataset of 200K images with 450K textual captions from a variety of news websites, blogs, and social media posts. The dataset and source code is publicly available at https://shivangi-aneja.github.io/projects/cosmos/.
Forward citations
Cited by 5 Pith papers
-
Verification-Notebook Learning for Source-Aware Multimodal Misinformation Detection
Verification-Notebook Learning distills labeled multimodal verification experience into a compact fixed notebook that lifts frozen-LVLM source-aware misinformation detection above prompting, cases, and agents.
-
XFacta: Contemporary, Real-World Dataset and Evaluation for Multimodal Misinformation Detection with Multimodal LLMs
XFacta is a new real-world, post-January-2024 multimodal misinformation dataset from X, and evaluations show that MLLM detectors need external evidence, especially image-to-text evidence, with multi-step reasoning per...
-
Dataset of News Articles with Provenance Metadata for Media Relevance Assessment
A new benchmark dataset and two tasks let researchers test whether AI systems can judge if a news image's recorded location and date match the article, with current chatbots scoring 64-81% on location but 42-58% on date.
-
Modeling Human Responses to Multimodal AI Content
A 154K-post study reports that humans identify AI content best when text and images are both present and inconsistent, and offers metrics plus an LLM agent for human-aligned responses.
-
Large Language Models and Provenance Metadata for Determining the Relevance of Images and Videos in News Stories
A prototype combines LLM reasoning with C2PA provenance metadata to classify news images and videos as relevant or not, without any benchmark evaluation.
Discussion (0). Sign in to comment.