GenImage
PulseAugur coverage of GenImage — every cluster mentioning GenImage across labs, papers, and developer communities, ranked by signal.
-
New DailyBench benchmark reveals AI image detectors struggle with modern fakes
Researchers have introduced DailyBench, a new benchmark designed to evaluate the performance of AI-generated image detectors against modern generative models and manipulation techniques. The benchmark includes two subse…
-
New deepfake detector adapts to evolving generative models
Researchers have developed BitMind Forensics (BMF), a novel deepfake detection system designed to continuously adapt to evolving generative models. Unlike static detectors that degrade in real-world performance, BMF is …
-
AI-generated image detector fragility exposed in new audit · 2 sources tracked
A new audit of training-free AI-generated image detectors reveals significant fragility and inconsistencies. The study found that implementation details, such as the choice of backbone network (e.g., AlexNet vs. VGG-16)…
-
ForensicConcept framework improves AI-generated image detection
Researchers have developed a new framework called ForensicConcept to improve the detection of AI-generated images. This method extracts explicit forensic concepts from existing detectors, making them transferable to dif…
-
New FiSeR method improves AI image detection across domains
Researchers have developed a new method called FiSeR to improve the detection of AI-generated images, particularly when faced with domain shifts. FiSeR employs a hierarchical contrastive learning framework that distingu…
-
New PGC framework enhances AI-generated image detection accuracy
Researchers have developed a new framework called Peak-Guided Calibration (PGC) to improve the detection of AI-generated images. This method focuses on aggregating salient, local features using a peak-sensitive mechanis…
-
New method uses image embedding sensitivity to detect AI-generated content
Researchers have developed a novel method for detecting AI-generated images by analyzing the sensitivity of intermediate layers in data embeddings. This approach examines the similarity between original and perturbed im…